Beruflich Dokumente
Kultur Dokumente
Computational Methods
and Experiments
WITeLibrary
Home of the Transactions of the Wessex Institute.
Papers presented at Materials Characterisation V are archived in the WIT eLibrary
in volume 72 of WIT Transactions on Engineering Sciences (ISSN 1743-3533).
The WIT eLibrary provides the international scientific community with immediate
and permanent access to individual papers presented at WIT conferences.
http://library.witpress.com
CONFERENCE CHAIRMEN
A.A. Mammoli
University of New Mexico, USA
C.A. Brebbia
Wessex Institute of Technology, UK
A. Klemm
Glasgow Caledonian University, UK
Organised by
Wessex Institute of Technology, UK
University of New Mexico, USA
Sponsored by
WIT Transactions on Engineering Sciences
WIT Transactions
Transactions Editor
Carlos Brebbia
Wessex Institute of Technology
Ashurst Lodge, Ashurst
Southampton SO40 7AA, UK
Email: carlos@wessex.ac.uk
Editorial Board
B Abersek University of Maribor, Slovenia
Y N Abousleiman University of Oklahoma,
USA
Spain
Belgium
Wales, Australia
UK
Germany
Dresden, Germany
Singapore
J M Hale University of Newcastle, UK
K Hameyer Katholieke Universiteit Leuven,
Belgium
C Hanke Danish Technical University,
Denmark
K Hayami University of Toyko, Japan
Y Hayashi Nagoya University, Japan
L Haydock Newage International Limited, UK
A H Hendrickx Free University of Brussels,
Belgium
C Herman John Hopkins University, USA
S Heslop University of Bristol, UK
I Hideaki Nagoya University, Japan
D A Hills University of Oxford, UK
W F Huebner Southwest Research Institute,
USA
J A C Humphrey Bucknell University, USA
M Y Hussaini Florida State University, USA
W Hutchinson Edith Cowan University,
Australia
T H Hyde University of Nottingham, UK
M Iguchi Science University of Tokyo, Japan
D B Ingham University of Leeds, UK
L Int Panis VITO Expertisecentrum IMS,
Belgium
N Ishikawa National Defence Academy, Japan
J Jaafar UiTm, Malaysia
W Jager Technical University of Dresden,
Germany
Y Jaluria Rutgers University, USA
C M Jefferson University of the West of
England, UK
P R Johnston Griffith University, Australia
D R H Jones University of Cambridge, UK
N Jones University of Liverpool, UK
D Kaliampakos National Technical
University of Athens, Greece
N Kamiya Nagoya University, Japan
D L Karabalis University of Patras, Greece
Thessaloniki, Greece
Technology, USA
Champaign, USA
B Ribas Spanish National Centre for
Environmental Health, Spain
K Richter Graz University of Technology,
Austria
S Rinaldi Politecnico di Milano, Italy
F Robuste Universitat Politecnica de
Catalunya, Spain
J Roddick Flinders University, Australia
A C Rodrigues Universidade Nova de Lisboa,
Portugal
F Rodrigues Poly Institute of Porto, Portugal
C W Roeder University of Washington, USA
J M Roesset Texas A & M University, USA
W Roetzel Universitaet der Bundeswehr
Hamburg, Germany
V Roje University of Split, Croatia
R Rosset Laboratoire dAerologie, France
J L Rubio Centro de Investigaciones sobre
Desertificacion, Spain
T J Rudolphi Iowa State University, USA
S Russenchuck Magnet Group, Switzerland
H Ryssel Fraunhofer Institut Integrierte
Schaltungen, Germany
S G Saad American University in Cairo, Egypt
M Saiidi University of Nevada-Reno, USA
R San Jose Technical University of Madrid,
Spain
F J Sanchez-Sesma Instituto Mexicano del
Petroleo, Mexico
B Sarler Nova Gorica Polytechnic, Slovenia
S A Savidis Technische Universitat Berlin,
Germany
A Savini Universita de Pavia, Italy
G Schmid Ruhr-Universitat Bochum, Germany
R Schmidt RWTH Aachen, Germany
B Scholtes Universitaet of Kassel, Germany
W Schreiber University of Alabama, USA
A P S Selvadurai McGill University, Canada
J J Sendra University of Seville, Spain
J J Sharp Memorial University of
Newfoundland, Canada
Q Shen Massachusetts Institute of Technology,
USA
X Shixiong Fudan University, China
G C Sih Lehigh University, USA
L C Simoes University of Coimbra, Portugal
Leuven, Belgium
Netherlands
Greece
Alicante, Spain
Germany
Hungary
Materials Characterisation V
Computational Methods
and Experiments
Editors
A.A. Mammoli
University of New Mexico, USA
C.A. Brebbia
Wessex Institute of Technology, UK
A. Klemm
Glasgow Caledonian University, UK
Editors:
A.A. Mammoli
University of New Mexico, USA
C.A. Brebbia
Wessex Institute of Technology, UK
A. Klemm
Glasgow Caledonian University, UK
Published by
WIT Press
Ashurst Lodge, Ashurst, Southampton, SO40 7AA, UK
Tel: 44 (0) 238 029 3223; Fax: 44 (0) 238 029 2853
E-Mail: witpress@witpress.com
http://www.witpress.com
For USA, Canada and Mexico
Computational Mechanics Inc
25 Bridge Street, Billerica, MA 01821, USA
Tel: 978 667 5841; Fax: 978 667 7582
E-Mail: infousa@witpress.com
http://www.witpress.com
British Library Cataloguing-in-Publication Data
A Catalogue record for this book is available
from the British Library
ISBN: 978-1-84564-538-0
ISSN: 1746-4471 (print)
ISSN: 1743-3533 (on-line)
The texts of the papers in this volume were set individually by the authors or under their
supervision. Only minor corrections to the text may have been carried out by the
publisher.
No responsibility is assumed by the Publisher, the Editors and Authors for any injury and/
or damage to persons or property as a matter of products liability, negligence or
otherwise, or from any use or operation of any methods, products, instructions or ideas
contained in the material herein.
WIT Press 2011
Printed in Great Britain by Martins the Printers.
All rights reserved. No part of this publication may be reproduced, stored in a retrieval
system, or transmitted in any form or by any means, electronic, mechanical, photocopying,
recording, or otherwise, without the prior written permission of the Publisher.
Preface
The increasing demands for high-quality products from both industry and consumers
are the driving force for the rapid developments in materials science and engineering.
In principle materials science involves relating the desired properties and relative
performance of a material to its microstructural features through characterization.
The major determinants of the structure of a material and hence its properties are
its constituent chemical elements and the way a material has been processed into
its final form.
Over the years, a variety of experimental techniques have been developed for
characterizing the physical and chemical properties of materials. Unfortunately
due to a number of simplifying assumptions and limitations on the use of individual
methods, it is not often possible to describe in a qualitative, reliable way the
microstructural features of many materials. Triangulation of different experimental
methods as well as computer simulations may become essential to achieve a
thorough, comprehensive analysis. Simulations can contribute to the understanding
of the phenomena and to provide a good basis for the development of durable
materials and components which can withstand ambient and extreme environmental
conditions.
The way forward in material characterisation is to develop new experimental
techniques or apply existing methodologies adopted from other related disciplines.
A very wide range of materials, starting with metals through polymers,
semiconductors to composites, necessitates a whole spectrum of experimental
techniques and numerical models, which are specific for material types. Some of
these well established methodologies could potentially find applications in new
fields. In this context a multidisciplinary approach in material characterisation and
the exchange of original ideas is indispensible.
The aim of the International Conference on Computational Methods and
Experiments in Materials Characterisation held in Kos, Greece, was therefore to
facilitate such interdisciplinary interactions within the research community. The
resulting conference book has been arranged in several chapters addressing various
experimental and numerical methods. The wide range of topics covers mechanical
characterisation and testing, corrosion problems and thermal analysis as well as
recycled materials, nano-composites and energy materials.
The editors would like to express their gratitude to all authors without whose
involvement this book could not have been produced. We wish to aknowledge the
valuable input of the members of the Scientific Advisory Committee in attracting
and selecting many high quality contributions. We trust that this book will present
some innovative ideas and will facilitate further developments in materials science.
The Editors,
Kos, Greece 2011
Contents
Section 1: Micro and nano characterisation of cementitious materials
(Special session organised by A. J. Klemm)
Application of positron annihilation lifetime spectroscopy to
nano-characterisation of polymer-modified mortars
P. Guagliardo, A. J. Klemm, S. N. Samarin & J. F. Williams ............................. 3
Multi-technique investigation of calcium hydroxide crystals at the
concrete surface
E. Gueit, E. Darque-Ceretti, P. Tintillier & M. Horgnies ................................. 15
Characterization of the influence of the casting mould on the
surface properties of concrete and on the adhesion of a protective coating
M. Horgnies, P. Willieme, O. Gabet, S. Lombard & M. Dykman...................... 27
Section 2: Nano-materials
HRTEM techniques applied to nanocrystal modeling: towards an
atom-by-atom description
D. G. Stroppa, L. A. Montoro, E. R. Leite & A. J. Ramirez ............................... 41
Ca(OH)2 nanoparticle characterization: microscopic investigation of
their application on natural stones
V. Daniele & G. Taglieri ................................................................................... 55
Nanocarbon composite materials with optical response on
radioactive waste
M. Vantsyan, G. Popova, E. Karpuzova, M. Bobrov, O. Plaksin
& E. Dabek ........................................................................................................ 67
Section 1
Micro and nano
characterisation of
cementitious materials
(Special session organised
by A. J. Klemm)
Materials Characterisation V
Abstract
Positron annihilation lifetime spectroscopy (PALS) has been applied to study the
microstructural features of immature cement mortars. Two types of cement
mortars containing superabsorbent polymers (SAPs) were studied, in addition to
Ordinary Portland Cement (OPC). The ortho-positronium lifetimes for all
samples were in the range of 1.70-1.73 ns, values that are close to that of free
water (1.7 ns) and hence suggest the presence of water-filled pores. Periodic
lifetime measurements showed that the intensity of this component decreased
slightly over a period of four weeks, indicating water loss associated with the
curing process, evaporation or a combination of the two.
Keywords: cement mortar, superabsorbent polymer, positron annihilation
lifetime spectroscopy, positronium, porosity, hydration, curing.
1 Introduction
Cement is a material of immense practical importance, but in spite of its almost
ubiquitous use, its microstructural characterisation still proves to be problematic.
The complexity of the problem is enhanced when the composition of cement is
modified through the use of auxiliary agents. Ordinary Portland cement is a
combination of gypsum (CS*H2) and clinker. Gypsum acts to prevent rapid
setting of the cement paste while clinker is the hydraulic binder. The major
constituents of clinker are tricalcium silicate (C3S), dicalcium silicate (C2S),
WIT Transactions on Engineering Sciences, Vol 72, 2011 WIT Press
www.witpress.com, ISSN 1743-3533 (on-line)
doi:10.2495/MC110011
4 Materials Characterisation V
tricalcium aluminate and tetracalcium alumina-ferrite (C4AF); C = CaO, S =
SiO2, A = Al2O3, F = Fe2O3, S* = SO3 and H = H2O (abbreviations commonly
used in industrial nomenclature) [1, 2].
The presence of water among the active compounds of cement leads to a
rearrangement of their structure and the initiation of the hardening process. The
initial hardening is caused by the hydration of C3S, which forms a gel-like
silicate and calcium hydrate phase referred to as C-S-H gel. These particles
ultimately crystallise and bind together the particles of sand or stone into a hard
mass. Other hydrates of the complexes described above are also formed,
although the reaction rates may differ considerably. The final product is a
hardened cement paste which is a mixture of unreacted cement particles,
hydration products and pores. Two types of pores are usually distinguished capillary pores, which comprise the water- or air-filled spaces between the
hydrates, with sizes in the range of 10-1000 nm, and smaller pores referred to as
gel pores, which are contained in the amorphous hydrate phase. Gel and capillary
pores form a continuous network of pores throughout the material [3, 4].
The mechanical properties of cementitious materials are heavily influenced by
porosity, with the volume and size distribution of pores controlling both the
strength and durability of the material [5]. A detailed characterisation of the pore
structure and the factors that affect it are thus crucial to advancing the design of
these materials; however, classical porosimetry methods such as mercury
intrusion porosimetry (MIP) and gas adsorption cannot always be relied upon to
extract information on pores smaller than a few nanometres [6]. In this regard,
positron annihilation lifetime spectroscopy (PALS) has considerable advantages
over these classical methods. PALS is sensitive to both open and closed porosity,
for example, whereas isolated pores are invisible to the aforementioned
techniques. In addition, PALS is sensitive to pores in the size range of 0.3-30 nm
(that is, in the size range of gel pores), and both pore sizes and relative
concentrations can be measured (refer to the following section for a detailed
explanation) [7]. In this work, PALS was used to study the features of the
porosity of three cement mortars. Two types of cement mortars containing
superabsorbent polymers (SAPs) were studied, in addition to a sample composed
only of OPC.
SAPs are cross-linked networks of hydrophilic polymers with a high capacity
for water uptake - they can absorb and retain up to 500 times their weight in
water. This makes them ideal for use in water-absorbing applications such as
absorbent medical dressings and controlled release media [8]. During the mixing
process the polymers absorb the pore solution immediately after their addition to
the mortar, reaching saturation within minutes [9, 10]. They then swell to form
spherical cavities filled with water. At later stages of the hydration process the
water is released to the concrete matrix and the cavities remain as empty pores.
For mortars with a low water-to-cement ratio, it is possible to replace part of the
irregular capillary pores with larger spherical pores formed by saturated
polymers. The dispersion and size of these pores can be estimated by the
material attributes of polymers.
Materials Characterisation V
The samples were studied with PALS over a period of 4 weeks with
measurements starting when the samples were aged 21 days, in order to elucidate
the changes in nano-porosity caused by the addition of SAPs as well as the
changes arising from the curing/hydration process. In addition, samples aged at 3
and 6 weeks were studied with MIP to determine changes in their pore
distributions and bulk densities. MIP can provide information on larger pores
(from 100 nm to 100 m) to which PALS is not sensitive.
(1)
6 Materials Characterisation V
annihilation, the o-Ps lifetime in matter is significantly shorter than its vacuum
value (and can be as low as ~ 1 ns) [7].
The basis for studying porosity is that o-Ps tends to localise in regions of low
density, such as pores. This is because in the bulk of the material the positron
and the electron in o-Ps experience repulsive Coulomb and exchange forces
respectively. As a result, most of these materials (porous insulators) have a
negative work function for positronium [7]. Porosimetry is possible because the
annihilation rate (the inverse of the lifetime) in the pore is a function of the pore
size smaller pores have a shorter lifetime compared to large pores, where the
overlap of the Ps wave function with the pore walls is reduced. A semi-empirical
quantum mechanical relation has been developed by Tao [14] and Eldrup et al.
[15] to relate the o-Ps lifetimes to pore radii:
1
= 2 1-
R
R+R
1
2
sin
2R
(2)
R+R
3 Experimental
3.1 Materials and mixes
For the purpose of this research Portland cement (BS EN 197-1 CEM II/B-V
32,5) was mixed at 1:1 ratio with fine sand (the vast majority of particles were
distributed below 0.6 mm). Throughout the investigation the total water-tocement ratio of 0.45 was maintained. The mix compositions are presented in
Table 1. The pastes were shaped into cylinders with diameters of about 25 mm
and thicknesses in the range of 5-10 mm.
Table 1:
Sample designation
OPC
SAP-A
SAP-B
Mix code
(Water/Cement)total
0.45
0.45
0.45
(Water/Cement)effective
0.45
0.425
0.438
Sand/Cement
0.25
0.25
Materials Characterisation V
The SAPs used in this study were cross-linked polymers provided by BASF.
SAP-A is a copolymer of acrylamide and acrylic acid and SAP-B is a polymer
based on acrylic acid. The products had absorption capacities of 200-250 ml/g in
demineralised water, though the absorption in mortar depended on the product
and was approximately 10 mL/g for SAP-A and 5 mL/g for SAP-B. Both
materials were prepared by grinding and screening to sizes of 63-125 m, but
there was also a minor (less than 10%) content of finer particles. The SAP
particle shapes were irregular and their sizes after initial absorption were in the
range of 135-270 m for SAP-A and 105-210 m for SAP-B.
3.2 Positron lifetime and mercury intrusion measurements
For positron lifetime measurements, approximately 30 Ci of 22NaCl was
deposited on 7m Kapton foil and covered with an identical foil. The edges of
this foil-sandwich were then sealed, and this source foil was placed between two
identical pieces of sample. The positron lifetimes were measured with a fast-fast
coincidence system. The gamma-ray detectors consist of a truncated cone (31.8
mm diameter tapering to 12.7 mm with a height of 12.7 mm) BC418 scintillator
coupled to a Burle 8850 photomultiplier tube. The time resolution of the system
is approximately 220 ps, as determined from analysis of a spectrum of highpurity annealed nickel. The spectra comprise of at least 2 million counts and
have been analysed using PAScual version 1.3.0 [18].
Mercury intrusion porosimetry (MIP) has been carried out with the use of a
Porosimeter Autopore IV 9500 by Micromeritics, with a pressure range up to
60000 psi for all samples at ages 3 and 6 weeks.
8 Materials Characterisation V
nanosecond components are associated with annihilation in water contained in
capillary pores of the cement and hence cannot be directly related to a pore size.
Consolati and Quasso [20, 21], Consolati et al. [22] and Salgueiro et al. [23]
observed similar lifetimes in Portland cement pastes and attributed this
component to water-filled pores.
OPC
SAP-B
SAP-A
Table 2:
Age
(weeks)
3
4
5
6
3
4
5
6
3
4
5
6
2 (ps)
218 (9)
227 (9)
242 (7)
233 (8)
219 (5)
229 (5)
221 (6)
237 (6)
224 (6)
239 (5)
238 (5)
212 (9)
410 (10)
410 (10)
430 (10)
420 (10)
409 (7)
421 (9)
409 (8)
420 (10)
402 (8)
430 (10
420 (10)
390 (10)
3 (ns)
I1 (%)
I2 (%)
I3 (%)
1.67 (0.02)
1.67 (0.03)
1.73 (0.03)
1.72 (0.03)
1.73 (0.02)
1.76 (0.03)
1.71 (0.03)
1.63 (0.03)
1.71 (0.02)
1.67 (0.02)
1.67 (0.03)
1.73 (0.03)
43 (4)
46 (5)
55 (4)
50 (4)
45 (2)
49 (3)
46 (3)
53 (4)
47 (3)
57 (3)
53 (3)
43 (3)
50 (3)
48 (5)
40 (4)
44 (4)
49 (2)
46 (3)
50 (3)
43 (4)
48 (3)
39 (2)
42 (3)
53 (4)
5.9 (0.1)
5.7 (0.2)
5.2 (0.2)
5.2 (0.2)
5.0 (0.1)
4.0 (0.1)
4.3 (0.1)
4.2 (0.2)
4.6 (0.1)
4.3 (0.1)
4.2 (0.1)
4.3 (0.1)
0.93
1.04
0.99
0.99
1.15
1.08
1.04
1.08
1.01
1.01
1.09
1.07
In Figure 1, o-Ps lifetimes (3) and intensities (I3) are plotted as a function of
sample age. Lifetime spectra were recorded every 7 days for a 4 week period.
Over the course of the measurements, I3 for the cement sample containing SAPA is somewhat higher than for the cement containing SAP-B and the pure
cement sample. If these lifetimes are indicative of water-filled pores then this
suggests a higher concentration of pores in the sample containing SAP-A. As the
results for SAP-B are closer to those of pure cement, this suggests that the
presence of the polymer does not significantly alter the pore concentration.
For the SAP-A sample, a gradual decrease in I3 is observed over weeks 3-5,
whereas for the SAP-B and pure cement samples there is a decrease after the
initial measurement (weeks 3-4), and then I3 is relatively constant. While these
changes are small, they may be indicative of various stages of the hydration
process, as the decrease in I3 is consistent with the reduction in total porosity that
is known to occur as hydration progresses [3]. The gradual decline seen in the
SAP-A sample suggests that this polymer releases its water more slowly
compared to SAP-B.
At the onset of the measurement (age 21 days) the hydration process is likely
to be still underway, with both di- and tri-calcium silicates reacting with water to
produce a network of calcium silicate hydrates. Water could then be consumed in
the hydration process for at least the next two weeks. However, during this time
it is difficult to say if the loss of water seen by the decrease in I3 is also
associated with evaporation. As the samples lose water one might expect the
lifetimes to increase as there would be a transition from filled to empty pores.
Materials Characterisation V
The fact that this does not occur suggests that pores collapse with water loss (the
volume of capillary pores is known to decrease with hydration [3]).
These results are consistent with the work of Consolati and Quasso [20, 21]
who observed a systematic decrease in I3 with sample age, but they are in
contrast to the work of Myllyl and Karras [24] where the opposite trend was
seen. It should be noted however that a direct comparison with these studies is
not possible due to differing experimental conditions (namely the composition
and age of samples).
6.5
6.0
5.5
5.0
4.5
4.0
3.5
3
2.2
2.0
1.8
1.6
1.4
1.2
3
o-Ps intensities (top) and lifetimes (bottom) vs. sample age for
SAP-A (circles), SAP-B (triangles) and Ordinary Portland Cement
(squares).
10 Materials Characterisation V
0.25
0.2
0.15
0.1
0.05
0
1000000 100000
10000
1000
100
10
in the size range of 10-100 nm. The MIP results are in qualitative agreement with
the PALS data in the sense that the sample containing SAP-B displays similar
behaviour to pure cement - at 6 weeks their pore distributions almost coincide
and the total porosity for both samples has decreased with aging. This decrease is
also consistent with the reduction in I3 seen in the PALS data.
0.25
0.2
0.15
0.1
0.05
0
1000000 100000
10000
1000
100
10
Materials Characterisation V
11
the presence of SAP-A significantly alters the hydration process of the cement
mortar.
The bulk densities of the samples were also determined by MIP
measurements (Table 3). It was found that the density of all samples increased
with aging time. Again, the changes in density for the pure cement and SAP-B
samples are similar. The change in density for SAP-A is much smaller, which is
consistent with the slower rate of change of I3.
Table 3:
Sample
OPC
SAP-A
SAP-B
In order to gain further insight into the nature of the porosity, the samples
were heated at 116C for 4 hours under vacuum conditions (10-5 Torr) and
lifetime measurements were repeated. Lifetime spectra were recorded directly
after heating and the samples were sealed in a plastic bag containing a desiccant
(to prevent water absorption) during the measurement. The resulting lifetimes
and intensities are given in Table 4 and the corresponding pore sizes are given in
Table 5.
At least three types of water can be distinguished in a hydrating cement paste.
Chemically-bound water is directly incorporated into the structure of the
hydration products, physically-bound water is absorbed on the surfaces of
cement particles and reaction products, and there is also free water contained
in the capillary and gel pores [25]. The heat treatment will have removed a
significant fraction of the latter two sources of water; however, much higher
temperatures (above 950C) are required to remove chemically-bound water.
Table 4:
1 (ps)
2 (ps)
3 (ns)
4 (ns)
I1 (%)
I2 (%)
I3 (%)
I4 (%)
2
OPC
263 (2)
536 (7)
4.3 (0.1)
75 (2)
24 (1)
1.3 (0.1)
1.13
Sample
SAP-A
214 (9)
380 (9)
1.32 (0.07)
12.6 (0.7)
36 (8)
60 (7)
3.3 (0.4)
1.0 (0.1)
1.05
SAP-B
219 (4)
413 (5)
1.62 (0.05)
14.9 (0.6)
44 (2)
52 (2)
2.5 (0.1)
0.70 (0.01)
1.01
12 Materials Characterisation V
After heat treatment, 3 for OPC increases to 4.3 ns. This corresponds to a gel
pore diameter of 0.86 nm using the TE model (equation 2). However, the
intensity of this component decreases by a factor of about 3.3; this could suggest
that a significant fraction of the pores collapse upon water removal. There is also
a significant redistribution of intensity from I2 to I1 after heat treatment. These are
the shorter components associated with multiple positron states in the material.
The redistribution of intensity from the longer component (2) to the shorter one
(1) indicates that the heat treatment has reduced the number of positron traps in
the bulk material. This could be associated with an increase in the bulk density
after heating.
Table 5:
Pore diameters, d3 and d4, calculated from o-Ps lifetimes, and total
o-Ps intensities (Io-Ps).
d3 (nm)
d4 (nm)
Io-Ps before (%)
Io-Ps after (%)
OPC
0.86
4.3
1.3
Sample
SAP-A
0.42
1.45
5.2
4.3
SAP-B
0.50
1.56
4.2
3.2
For the cement samples containing SAPs an additional low intensity, long
lifetime component (4) appears in the spectra after heat treatment (and due to
the presence of this long component it was necessary to fix the background to the
left of the main coincidence peak to the average value in the analysis). 4 may be
associated with pores formed by the polymers which have replaced some of the
ordinary pore structure, and its emergence shows that the presence of SAP
significantly alters the structure of the cement paste.
For the sample containing SAP-A, 3 = 1.32 ns and 4 = 12.6 ns, with
respective intensities of 3.3% and 1.0%. These lifetimes would correspond to
pores having diameters of 0.42 and 1.45 nm respectively. There is also a
redistribution of the weighting of the shorter components (1 and 2); however
this is not as pronounced as for ordinary Portland cement.
For the sample containing SAP-B, 3 = 1.62 ns and 4 = 14.9 ns, with
intensities of 2.5% and 0.7% respectively. These lifetimes correspond to pores
having diameters of 0.42 and 1.45 nm respectively. The value of 3 is comparable
to that obtained before heat treatment (which was similar to the o-Ps lifetime in
water). This could suggest that the heat treatment did not result in the complete
removal of free water.
For both SAP-A and SAP-B the total o-Ps intensity (I3 + I4) decreases after
heat treatment, signalling a reduction in porosity; however, the reduction is
larger for the ordinary Portland cement sample.
Materials Characterisation V
13
5 Conclusion
This study has shown that PALS can be used to monitor the hydration process in
cement samples by using the o-Ps intensity as an indicator of the concentration
of water-filled pores. In addition, PALS has shown that the addition of a SAP to
the cement mixture alters the structure and hydration process of cement. The
addition of SAP-A caused the hydration process to progress more gradually
compared to OPC; however, the addition of SAP-B gave comparable results to
that of OPC.
Both PALS and MIP showed that the presence of SAP-A has a significant
effect on the hydration process. In addition, MIP showed that the concentration
of pores in the 10-100 nm range increases with aging for the SAP-A sample.
For the samples with SAP added, heat treatment under evacuation (resulting
in the removal of free water) gave rise to an additional long lifetime component
and a higher overall o-Ps intensity compared to OPC. This shows that the SAP
had modified the pore distribution, replacing some of the regular pore
distribution with larger pores formed by saturated polymers.
References
[1]
[2]
[3]
[4]
[5]
[6]
[7]
[8]
[9]
[10]
[11]
[12]
[13]
[14]
14 Materials Characterisation V
[15] Eldrup, M., Lightbody, D. & Sherwood, J. N. The temperature dependence
of positron lifetimes in solid pivalic acid. Chemical Physics 63, 51-58,
(1981).
[16] Gidley, D. W. et al. Positronium annihilation in mesoporous thin films.
Physical Review B 60, R5157 (1999).
[17] Gidley, D. W. et al. Determination of pore-size distribution in lowdielectric thin films. Applied Physics Letters 76 (2000).
[18] Pascual-Izarra, C. et al. Advanced fitting algorithms for analysing positron
annihilation lifetime spectra. Nuclear Instruments and Methods in Physics
Research Section A: Accelerators, Spectrometers, Detectors and Associated
Equipment 603, 456-466, (2009).
[19] Patro, A. P. & Sen, P. Anomalous parapositronium lifetime in water at 20
degrees C. Journal of Physics C: Solid State Physics 5, 3273 (1972).
[20] Consolati, G. & Quasso, F. Evolution of porosity in a Portland cement paste
studied through positron annihilation lifetime spectroscopy. Radiation
Physics and Chemistry 68, 519-521 (2003).
[21] Consolati, G. & Quasso, F. A positron annihilation study on the hydration
of cement pastes. Materials Chemistry and Physics 101, 264-268 (2007).
[22] Consolati, G., Dotelli, G. & Quasso, F. Positron lifetime spectroscopy as a
probe of nanoporosity of cement-based materials. Radiation Physics and
Chemistry 58, 727-731 (2000).
[23] Salgueiro, W., Somoza, A., Cabrera, O. & Consolati, G. Porosity study on
free mineral addition cement paste. Cement and Concrete Research 34, 9197, (2004).
[24] Myllyl, R. & Karras, M. Positron Annihilation Probing for the Hydratation
Rate of Cement Paste. Applied Physics 7, 303-306 (1975).
[25] Sen Wang, P. U. et al. 1H nuclear magnetic resonance characterization of
Portland cement: molecular diffusion of water studied by spin relaxation
and relaxation time-weighted imaging. Journal of Material Science 33,
3065-3071 (1998).
Materials Characterisation V
15
Abstract
The durability and aesthetic qualities of high-performance concrete, which
makes it particularly suitable for architectural applications, are constantly
compromised by environmental aggressions. In this study, an innovative solution
was developed to protect the concrete from these aggressions, which consists of
growing a mineral coating on the concrete surface. The coating is composed of
layered calcium hydroxide crystals, whose nucleation and growth are triggered
byvarious non-ionic surfactants (the details of the process will not be presented).
This paper describes the procedure used to investigate the structure of the formed
crystals. Scanning Electron Microscopy and optical microtopography were used
to determine the morphology of the crystals. Image analysis allowed the
quantification of their amount, size and shape. The contribution and limits of
each technique are discussed.
Keywords: concrete, scanning electron microscopy, image analysis, optical
microtopography, surface.
1 Introduction
The mechanical and aesthetic durability of concrete is often compromised by the
constant environmental aggressions to which the structures are exposed (organic
or inorganic particles, algae, micro-organisms, staining from various sources). It
is possible to protect concrete from these attacks and increase its durability by
applying organic coatings on the hardened surface, but this comes with
operational and environmental costs.
16 Materials Characterisation V
An innovative solution was proposed [1], which consists on covering and
protecting high-performance concrete (HPC) with a mineral coating made of
calcium hydroxide crystals (CH). The crystals growth happens during the
concrete setting and is triggered by the presence of non-ionic surfactants at the
concrete/formwork interface. One of the difficulties of this study was to
determine the adequate characterization methods to properly investigate the
influence of various surfactants on the amount, morphology and size of the
crystals. Several techniques exist to observe, measure and study concrete
hydrates, but not all of them are directly suitable for surface investigation.
An original procedure had to be developed, combining several techniques.
The structure and morphology of the crystals were assessed through Scanning
Electron Microscopy (SEM) both on the concrete surface and on polished
section. The SEM observations were completed by microtopography on the
concrete surface. The quantification of the amount and size of the crystals was
made by image analysis on binocular images of the concrete surface. The
purpose of this paper is to illustrate what each of these techniques can bring to
concrete surface studies.
Figure 1:
Materials Characterisation V
17
Figure 1 shows observations of both the surface (A) and the cross-section (B)
of a standard high-performance concrete. No specific features are visible on the
surface, except for a couple of scratch due to the formwork defects. As for the
polished section, it reveals a classical concrete microstructure, with dark grey
aggregates and light grey un-hydrated cement grains surrounded by cement
paste.
Figure 2 shows SEM observation of two cross-sections from two different
samples. These observations reveal the presence at the concrete surface of a thin
layered structure presenting various orientations and organization, sometimes
well-aligned parallel to the surface (A), sometimes arranged in a more chaotic
way in the first micrometers of the surface (B). At this level of observation, it is
not possible to clearly indentify the nature of this unexpected phase. The limited
magnification in back-scattered mode, the thin structure of the hydrates and the
surface damages due to the polishing all complicate the interpretation. This is
why it is necessary to make complementary observations of the surface itself.
Figure 2:
Figure 3 shows the same samples as figure 2 observed from the surface in
secondary electron mode. These micrographs allow a better understanding of the
structure of the crystals. All of them share the same layered and flaky structure
with different orientations. In all cases, the crystals are composed of thin leaves
that grow around the nucleation point. For some crystals, these leaves are strictly
parallel to the surface and grow as flower petals around the central point,
reaching sometimes a perfect hexagonal shape. For other crystals, the leaves are
strictly perpendicular to the surface, forming a very regular spherulitic structure.
This very organized structure was not visible on the polished sections, where the
leaves appeared randomly implanted in the surface. This is due to the polishing,
which damaged the first micrometers of the surface and disturbed the structure.
The characteristic hexagonal shape of the crystals, as well as their layered
organization, allow identifying them as calcium hydroxide no other concrete
phase would present this morphology. This is confirmed by EDS analysis, where
mostly calcium is detected is these areas (spectrum not showed).
WIT Transactions on Engineering Sciences, Vol 72, 2011 WIT Press
www.witpress.com, ISSN 1743-3533 (on-line)
18 Materials Characterisation V
Figure 3:
Materials Characterisation V
19
Figure 4:
Figure 5:
20 Materials Characterisation V
The same measurements can be made of figure 5. The mapping was acquired
on the same sample, but in an area were the CH crystals have been pulled out of
the surface during the demoulding phase, leaving hollows the size of the crystals.
The depth of the hollows gives the height between the surface and the bottom of
the crystals.
The measurements from figures 4 and 5 give an estimation for the total
thickness of the crystals of 20 micrometers. Of course, the operation has to be
repeated a large number of times to obtain statistical results. This is only one
example taken from one of the samples.
It should be noted that profilometry gives no conclusive information for the
samples where the crystals are oriented perpendicular to the surface. Figure 6
shows a mapping of such a sample (A): the spherulitic structures appear slightly
lighter than the surface, but are difficult to distinguish. A zoom on the mapping
(B) reveals that the crystals are indeed apparent, but that the resolution is far
from being sufficient to separate the different flakes composing the structure.
Figure 6:
Materials Characterisation V
21
The CH crystals are shinier than the cement paste, which makes them visible
with the naked eye, but this is not sufficient for image analysis software to
distinguish them. To overcome this problem, the samples were coloured with a
black felt-tip pen prior to analysis. The black ink penetrates the cement paste,
strongly colouring it, but is not absorbed by the crystals which then become
clearly distinguishable.
For each concrete samples, five small areas of the surface (1 cm2) were
coloured and photographed under a binocular. The images were treated and
analysed using the open-source software ImageJ [7].
The most difficult step in image analysis is to separate the studied particles
from the background. In our case, the following procedure was used.
Figure 7 shows an example of an image prior to treatment. The brightness and
contrast of the images were optimized using the automatic function on the
software. The optimization was not based on the whole image, but on the
histogram analysis of a small area of one of the particle on the image. This step
created a highly contrasted image where particles appeared red and the
background black (figure 8).
Figure 7:
Figure 8:
22 Materials Characterisation V
The noise was removed using the remove outliers function of the software,
which replaces each pixel by the median value of its neighbours in a given area
(figure 9). In this case, a radius of 20 pixels was considered, and the removing
was set to occur if the difference between the considered pixel and the median of
its neighbour was higher than 10.
Figure 9:
The images were then converted to binary and inverted. At this point, the
background appeared white and the particles black (figure 10).
Figure 10:
The holes in the particles were closed using the Fill holes function of the
software (figure 11).
Finally, the watershed function was used to automatically draw the outlines
of the particles (figure 12). This step might be the main cause of errors in the
measurements, as watershed segmentation works best for convex objects that do
not overlap too much. In some cases, the software was not able to correctly
separate clusters of particles, which were then counted as one big particle.
Materials Characterisation V
Figure 11:
Figure 12:
23
Once the image was treated, the particles were counted one by one. The
smallest ones (with a surface smaller than 0.1 mm) were not considered as they
are more likely residual noise than actual crystals. The particles overlapping with
the boarder of the image were not considered either as their size would not be
correctly measured. This means than the measured amount of crystals (from 1 to
20 particles per mm depending on the samples) is slightly smaller than reality.
The following parameters were measured for each particle:
- The coordinates of the centre of the particles. These coordinates were
used to calculate the distance between each crystal and its closest
neighbour and to verify that the nucleation appeared randomly and
homogeneously on the surface (figure 13).
- The Ferets diameter, which is the longest distance between two
opposing points of the particle.
- The fraction area of the surface covered by the crystals. As these
crystals are artificially grown to act as a mineral coating, this fraction
area must be as high as possible. The results showed that even if the
crystals seemed very large, they hardly cover 50% of the surface, except
in one promising case where almost 100% of the surface was coated.
WIT Transactions on Engineering Sciences, Vol 72, 2011 WIT Press
www.witpress.com, ISSN 1743-3533 (on-line)
24 Materials Characterisation V
-
Figure 13:
Figure 14:
Materials Characterisation V
25
4 Conclusion
Three techniques scanning electron microscopy, microtopography and image
analysis were successfully used to assess the morphology and geometrical
parameters of calcium hydroxide crystals at the concrete surface. These
techniques are well-known and developed, but not necessarily widely used in
concrete research. Yet, they have proved very efficient in this case to make a
preliminary study of a new phenomenon the massive growth of CH crystals in
presence of surfactants, bringing complementary information and results. Of
course, they absolutely do not make further investigation any less necessary. For
example, a proper crystallographic study would be essential to fully understand
the growing mechanisms and the action of the surfactants.
In the field of materials characterization, it is important to be creative and to
combine and adapt existing techniques. This is particularly true in the field of
concrete research, and even more when it comes to concrete surface, a topic
which is slowly emerging and where a lot of fascinating research still waits to be
done.
References
[1] Gueit, E., Darque-Ceretti, E., Tintillier, P. & Horgnies, M., Surfactantinduced growth of calcium hydroxide at the concrete/formwork interface as
a mineral coating for concrete, Manuscript submitted for publication.
[2] Skalny, J., Gebauer, I. & Odler, I., (eds). Calcium Hydroxide in Concrete,
The American Ceramic Society: Westerville, 2001.
[3] Garbacz, A., Courard, L., & Kostana, K., Characterization of concrete
surface roughness and its relation to adhesion in repair systems, Materials
Characterization, 56, pp. 281-289, 2006.
[4] PER09 Perez, F., Bissonette, B. & Courard, L., Combination of mechanical
and optical profilometry techniques for concrete surface roughness
characterisation. Magazine of Concrete Research, 61(6), pp. 389-400, 2009.
[5] Ramirez, A.M., Demeestere, K., De Belie, N., Mntyl, T., & Levnen,
E, Titanium dioxide coated cementitious materials for air purifying
purposes: Preparation, characterization and toluene removal potential,
Building and Environment, 45, pp. 832-838, 2010.
26 Materials Characterisation V
[6] Gallucci, E. & Scrivener, K., Crystallisation of calcium hydroxide in early
model and ordinary cementitious systems, Cement andConcrete Reseach,37,
pp. 492-501, 2007.
[7] NIH, http:\\rsbweb.nih.gov/ij
Materials Characterisation V
27
Abstract
Protective coatings are deposited on concrete to improve aesthetics and to
prevent ageing. However, their adhesion on concrete depends on several
interlinked parameters. In this study, the surfaces of concrete are characterized
according to the process of casting and post-treatment used (sandblasting) by
using Scanning Electron Microscopy (SEM), Fourier Transformed-Infrared (FTIR) spectroscopy and profilometry. The surface properties are correlated to the
adhesion force of a polyurea (PU) coating. The development of a specific peel
test (a strengthened and porous membrane is introduced into the layer of liquid
coating before its crosslinking) ensures a reproducible debonding of the
coating/concrete system and allows measuring the fracture energy. Moreover, the
interface after debonding is analyzed by FT-IR to highlight the presence of
concrete/coating residues and to determine the locus of failure. Results underline
that the nature of casting mould influences the concrete surface and modifies the
adhesion of PU coating. The mould made of polyoxymethylene (POM) induces a
micro-tearing of the extreme surface of concrete during demoulding. By
increasing the roughness and the open porosity of the concrete surface, this
tearing enhances the adhesion of the coating. On the contrary, the smooth
concrete surface, induced by the use of a polyvinylchloride (PVC) mould,
decreases the anchorage of the coating. Finally, the sandblasting of the surface,
by increasing the roughness and the interface area, is an interesting treatment to
promote the adhesion of PU coating, whatever the mould used for the casting.
Keywords: concrete, coating, roughness, FT-IR, SEM, peel test, adhesion.
28 Materials Characterisation V
1 Introduction
The staining of concrete could occur due to its specific microstructure that
retains the liquid and dust particles. The deposition of a coating is then important
to close the surface porosity and protect concrete against acid rains, settlement of
algae and lichensetc [1, 2]. PU coatings are commonly used due to their high
resistance against chemical and mechanical aggressions [3, 4]. However, the
surface properties of concrete depend on several interlinked parameters as
chemical composition, intrinsic porosity and roughness. The aim of this study
concerns the influence of the casting process and post-treatment on the adhesion
between PU coating and concrete surface. The influences of the casting
conditions on the hardened concrete surface were characterized by SEM, FT-IR
and profilometry. Secondly, this study was undertaken to determine if the surface
properties of concrete could influence the adhesion of PU coating. Fracture
energies were measured by a specific 90-peel test. This method was retained
because it is appropriate to characterize the adhesion of thin films [58]. Some
publications have already showed the use of strengthened membrane or mesh
sheet that were incorporated into the bulk of soft material to characterize [9].
Concerning our system, the introduction of a polymer membrane into the bulk of
the coating was necessary to strengthen the system and measure a reproducible
adhesion of the PU coating. After the peel tests, the FT-IR analyses of the
debonded faces were undertaken to detect the residues of concrete or coating and
determine the loci of failure. FT-IR spectroscopy allows detecting organic
compounds of coatings [10, 11] and several components of concrete [12].
Materials Characterisation V
29
The isocyanates units reacted with water to produce a polyurea-based film [13].
The coatings were sprayed (120 g/m) with air pressure of 3 atmospheres on the
concrete surface. The drying period was about 3 days under ambient conditions
(25C, 50% relative humidity) before peeling.
2.2 Methods of characterization
2.2.1 Scanning electron microscopy (SEM)
Samples were characterized by using a high-resolution field-effect gun digital
scanning electron microscope (SEM FEG Quanta 400 from FEI Company; using
an accelerating voltage of 15 keV and a current intensity of 1 nA). Images of the
cross-sections were obtained after being polished.
2.2.2 Profilometry
The roughness of concrete samples was measured with a Surftest SJ-201 M
mechanical profilometer (Mitutoyo) in order to calculate the arithmetic mean of
the profile deviations from the mean line (Ra). The Ra value was obtained by
compiling the arithmetic mean of 5 profiles of 12.5 mm.
2.2.3 Fourier transform-infrared spectroscopy (FT-IR)
The FT-IR spectrometer Nicolet iS10 (Thermo Fisher Scientific Inc.) was
equipped with a deuterated triglycine sulfate (DTGS) detector and controlled by
OMNIC software. The Attenuated Total Reflexion (ATR) mode was mainly used
in this study. FT-IR (in ATR mode) characterized the sample over a thickness of
a few m. The sampling area analyzed was approximately 1 mm. The crystal
used was made of diamond and 16 scans were routinely recorded over the range
4,000-650 cm-1 with a spectral resolution of 4 cm-1. The background was
collected at ambient atmosphere before analyzing each sample. Spectra were
corrected with a linear baseline. No specific preparations of the samples of
concrete and coating were performed before FT-IR analyses: they were studied
just after demoulding or after debonding.
2.2.4 Specific peel tests of concrete/coating system
Peel test allows measuring the debonding force (F). According to the peel angle
() and the width of the adhesive coating (w), the fracture energy (G) could be
calculated according to [14]:
F
G
w
1 cos
In the specific case of 90 peel angle, G is equal to the peeling force (F)
divided by the width (w) of the adhesive coating. The coating was strengthened
by inserting a membrane. The membrane was firstly cut in polyamide 6.6 tissues
(Nytex from Dutscher SAS) and deposited on each concrete samples. The
coating was directly sprayed on this membrane in order to avoid any air bubble
at the interface. Finally, the edges of the membrane were cut to avoid any shear
during the debonding. The membrane was flexible but non-stretched under the
30 Materials Characterisation V
solicitations that occurred during the peel test. The dimensions of each
membrane were 200 mm long, 20 mm wide, 120 m thick, with an open porosity
of 50% and a mesh opening of 150 m. We observed by SEM (Figure 1) a crosssection of the concrete/coating interface in order to describe the system. No
mechanical step was used to initiate the crack at the interface. All the 90-peel
tests were performed with a 1000 N sensor, equipped with a specific mobile
table (as detailed by Figure 2). All the peel tests were performed under ambient
conditions (25C, 50% relative humidity) and by using a constant speed of peel
(0.2 mm/s).
Strengthened membrane
PU coating
(130 m thickness)
Concrete surface
Figure 1:
Figure 2:
Materials Characterisation V
31
3 Results
3.1 Analyses of the surfaces of reference materials
3.1.1 Concrete surfaces according to the nature of the mould
Table 1 details the IR bands of concrete (demoulded with PVC). Portlandite (at
3640 cm-1) and C-S-H*/silica (large band at 1078 cm-1) were well detected.
Moreover, three strong bands were also assigned to the presence of CaCO3. Their
detection confirmed previous studies [12]. Detection of CaCO3 could be due to
the carbonation process but also due to the use of limestone filler in concrete
mix. As presented by Table 1, concrete demoulded with POM (spectrum not
shown) underlines distinct IR bands at 935 and 900 cm-1 that were assigned to
alite [15] (a component of the cement before hydration). This FT-IR spectrum
showed also peaks at 1092 and 800 cm-1, assigned to Si-O bonds from silica
fumes [12]. These silica fumes and alite are usually present into the bulk of
concrete sample. Moreover, no IR band of the portlandite (Ca(OH)2) was
detected.
Table 1:
Concrete
(PVC mould)
O-H; Ca(OH)2
3640
C=O; CaCO3
1080-970
Assignment
Concrete
(POM mould)
PU
coating
Si-O; C3S
935; 900
Si-O; silica
797; 777
N-H; urea
3335
NCO; isocyanates
2270
C=O; urea
1690
The concrete samples were also studied by SEM. Concrete demoulded with
POM (Figures 3a and 3b) appears to be rougher than concrete demoulded with
PVC (Figures 3c and 3d). The flat topography observed after using PVC mould
contrasted with the heterogeneous surface obtained after using a POM mould.
These observations confirmed the measurements obtained by profilometry
(Table 2): the Ra values of concrete could vary according to the demoulding
process. Indeed, the results obtained by SEM suggest that a tearing of the
extreme surface of concrete could happen during the removing of the POM
mould. This hypothesis could explain why silica fumes were easily observed by
SEM (Figure 3b) and why alite was detected by FT-IR spectroscopy (Table 1).
32 Materials Characterisation V
(a)
(b)
Silica fumes
(d)
(c)
Figure 3:
Table 2:
Ra roughness (m)
Reference
After sandblasting
Materials Characterisation V
33
Silica fumes
Figure 4:
CaCO3
(b)
(a)
Silica
Figure 5:
Alite
34 Materials Characterisation V
polyamide 6.6. was also analyzed by FT-IR in order to assign the reference peaks
(amide units at 3300, 3070, 1630 and 1530 cm-1). However, these IR bands of the
polyamide were never detected at the interface after debonding.
3.2 Results of peel tests and FT-IR analyses of the interface after debonding
3.2.1 Influence of the mould used to cast the concrete
Figure 6 compares the peel curves obtained during the debonding of PU coating
from concrete samples. The adhesion of the coating varied according to the
nature of the casting mould.
1400
1200
(a) Concrete
(POM mould)
G (N/m)
1000
800
600
(b) Concrete
(PVC mould)
400
200
0
0
10
20
30
40
50
60
70
80
Figure 6:
As described by Table 3, the fracture energy measured during the peel of the
PU coating from concrete (POM mould) is higher (1000 N/m +/- 150) than the
one measured during the peel from concrete (PVC mould): 400 N/m +/-50.
Figure 7a presents the FT-IR spectrum recorded on concrete (POM mould) side
after peeling: almost all the IR bands were assigned to the residues of coating
that covered the entire surface. These results confirm that the high fracture
energy was linked to a debonding into the bulk of the coating. It could also be
deduced that the adhesion between concrete and coating was better than 1000
N/m. On the contrary, Figure 7b shows FT-IR spectrum of concrete (PVC
mould) side after peeling: compounds of concrete (such as CaCO3, silica and
portlandite) were mainly detected. Concerning the internal side of coating after
peeling from concrete (PVC mould) sample, all the IR bands (spectrum not
shown) were assigned to the own components of the coating. No FT-IR bands
could be assigned to concrete residues.
Materials Characterisation V
Table 3:
Sample
35
Fracture
energy (N/m)
Sandblasting
400 +/- 50
At concrete/PU interface
Sandblasting
Concrete/POM
Concrete/PVC
Figure 7:
FTIR spectra of concrete sides after peeling: (a) if POM mould; (b)
if PVC mould.
These data allow concluding that the debonding occurred at the interface
between concrete and coating. The fracture energies and FT-IR analyses of the
locus of failure could be correlated to the SEM observations of concrete surface
WIT Transactions on Engineering Sciences, Vol 72, 2011 WIT Press
www.witpress.com, ISSN 1743-3533 (on-line)
36 Materials Characterisation V
before the deposition of the coating. Figure 3 had highlighted that the surface of
concrete demoulded with POM was more heterogeneous than concrete
demoulded with PVC, which was smoother. We can deduce that the liquid
coating (before crosslinking) strongly diffused and anchored into the
heterogeneous concrete surface, demoulded with POM. On the contrary, the
smooth and close concrete demoulded with PVC did not allow a high mechanical
anchorage of the coating.
3.2.2 Influence of the sandblasting of concrete surface
The peels tests undertaken on the sandblasted concrete (PVC mould) highlighted
high fracture energies (about 1200 N/m) while the peel undertaken on reference
concrete (PVC mould) samples induced a low level of adhesion of the PU
coating (350-400 N/m), as described in Figure 8a. Concerning concrete
demoulded with POM, the fracture energies (presented in Figure 8b) did not
1600
(a)
1400
Sandblasted concrete
(PVC mould)
1200
G (N/m)
1000
800
600
Not-sandblasted concrete
(PVC mould)
400
200
0
0
10
20
30
40
50
60
70
80
1600
(b)
1400
Sandblasted concrete
(POM mould)
1200
G (N/m)
1000
Not-sandblasted concrete
(POM mould)
800
600
400
200
0
0
10
20
30
40
50
60
70
80
Figure 8:
Materials Characterisation V
37
evolve: they were as high after sandblasting as after demoulding. Thanks to these
results, we can conclude that a sandblasting of concrete demoulded with PVC
could be interesting to enhance the adhesion of PU coating. As described by
SEM images (Figure 4) and values of roughness (Table 2), the open porosity and
the roughness of concrete demoulded with PVC were higher after the
sandblasting (Figure 4c) than after demoulding (Figure 3c). On the contrary, the
sandblasting was less helpful with concrete demoulded with POM because the
open porosity was already present after demoulding. FT-IR analyses of concrete
sides were performed after peeling but only residues of coating were detected on
these concrete sides (whatever the nature of mould used). The debonding
occurred inside the layer of the PU coating if the substrate had been sandblasted
before the deposition of the coating.
4 Conclusion
A specific methodology of 90-peel test was used to measure the fracture energy
between concrete and PU coating. By introducing a porous and thin membrane to
strengthen the layer of coating, measurements were reproducible and allowed
comparing the fracture energies with the FT-IR analyses of the loci of failure
recorded after debonding. These results were correlated to the surface properties
of the reference concrete surface after demoulding (topography, composition,
roughness) that were characterized by several methods. The influence of the
nature of mould was significant. Some moulds, such as POM, could induce a
micro-tearing of the extreme surface of concrete during the demoulding. This
phenomenon increased the roughness (and the open porosity) of concrete and
improved the anchorage of the liquid coating (before its crosslinking). On the
contrary, the smooth and flat concrete surface induced after using a PVC mould
did not allow a good adhesion of PU coating. The influence of the mechanical
anchorage was confirmed by the high fracture energies measured on the
sandblasted samples of concrete. In future, analyses of the surface porosity and
sorptivity will be performed to confirm these results.
References
[1] Dubosc, A., Escadeillas, G. & Blanc, P.J., Characterization of biolical
stains on external concrete walls and influence of concrete ad underlying
material. Cement and Concrete Research, 31, pp. 1613-1617, 2001.
[2] Manoudis, P.N., Karapanagiotis, I., Tsakalof, A., Zuburtikutis,
I., Kolinkeova, B. & Panayiotou, C., Surface properties of
superhydrophobic coatings for stone protection. Journal of Nano Research,
8, pp. 23-33, 2009.
[3] Chattopadhyay, D.K. & Raju, K.V.S.N., Structural engineering of
polyurethane coatings for high performance applications. Progress in
Polymer Science, 32, pp. 352-418, 2007.
38 Materials Characterisation V
[4] Carmona-Quiroga, P.M., Martnez-Ramrez, S., Sobrados, I. & BlancoVarela, M.T., Interaction between two anti-graffiti treatments and cement
mortar (paste). Cement and Concrete Research, 40, pp. 723-740, 2010.
[5] Barquins, M. & Ciccotti, M., On the kinetics of peeling of an adhesive tape
under a constant imposed load. International Journal Adhesion and
Adhesives, 17, pp. 65-68, 1997.
[6] Horgnies, M., Darque-Ceretti, E. & Combarieu, R., Adhesion of pressure
sensitive adhesives to automotive coatings: Influence of topcoat
composition. Journal of Adhesion Science and Technology, 18, pp. 10471061, 2004.
[7] Johnson, K.L., Kendall, K. & Roberts, A.D., Surface Energy and the
contact of elastic solids. Proceedings of Royal Society of London, 324, pp.
301-313, 1971.
[8] Kinloch, A.J., Adhesion and adhesives: science and technology. Chapman
and Hall, London, New York, pp. 66-73, 1987.
[9] Giannis, S., Adams, R.D., Clark, L.J. & Taylor, M.A., The use of a
modified peel specimen to assess the peel resistance of aircraft fuel tank
sealants. International Journal of Adhesion and Adhesives, 28, pp. 158-175,
2008.
[10] Almeida, E., Balmayore, M. & Santos, T., Some relevant aspects of the use
of FT-IR associated techniques in the study of surfaces and coatings.
Progress in Organic Coatings, 44, pp. 233-242, 2002.
[11] Poliskie, M. & Clevenger, J.O., Fourier Transform Infrared (FT-IR)
spectroscopy for coating characterization and failure analysis. Organic
Finishing, pp. 44-47, 2008.
[12] Chollet, M., Horgnies, M., Analyses of the surfaces of concrete by Raman
and FT-IR spectroscopies: comparative study of hardened samples after
demoulding and after organic post-treatment. Surface and Interface
Analysis, In press.
[13] Agrawal, R.K. & Drzal, L.T., Adhesion mechanisms of polyurethanes to
glass surfaces. Part I. Structure property relationships in polyurethanes and
their effects on adhesion to glass. Journal of Adhesion, 54, pp. 79-102,
1995.
[14] Kendall, K., Thin-film peeling the elastic term. Journal of Physics D:
Applied Physics, 8, pp. 1449-1452, 1975.
[15] Frost, R.L., Cejka, J. & Weier, M.L., Molecular structure of the uranyl
silicates a Raman spectroscopic study. Journal of Raman spectroscopy,
37, pp. 538-551, 2005.
[16] Daniel-da-Silva, A.L., Bordado, J.C.M. & Martin-Martinez, J.M., Moisture
curing kinetics of isocyanate ended urethane quasi-prepolymers monitored
by IR spectroscopy and DSC. Journal of Applied Polymer Science, 107, pp.
700-709, 2007.
[17] Agrawal, K. & Drzal, L., Adhesion mechanisms of polyurethanes to glass
surfaces. Journal of Adhesion Sciences Technology, 9, pp. 1381-1400,
1995.
Section 2
Nano-materials
Materials Characterisation V
41
Abstract
The development of technologies based on nanostructures presents a wide range
of challenges for materials scientists and engineers, including the attainment of
well-controlled synthesis procedures, the improvement of characterization
techniques down to the atomic scale resolution, and the conception and
validation of reliable models that can describe materials properties as functions
of their morphology and fabrication process. A relevant topic in this scenario is
the correlation among the spatial distribution of chemical elements, the surface
energy configuration, the growth mechanism, and the resultant nanocrystal 3D
morphology.
This work presents an overview on the use of advanced HRTEM techniques
for the quantitative analysis of nanocrystals and how these results can be used to
implement nanocrystals models, which can analytically describe the material
features on an atomic level. The presented findings show the combined use of
experimental data and theoretical tools, such as image simulation and ab initio
surface energy calculations, for the advanced quantitative characterization of
nanocrystalline systems.
The combination of experimental and theoretical efforts on HRTEM
characterization represents a powerful tool for the nanocrystal 3D morphology
elucidation with atomic resolution and the chemical/structural properties
assessment in a quantitative way. Thereby, it is presented as the stepping stone
towards the development of novel approaches to describe nanostructured
systems.
Keywords: HRTEM, nanocrystals modeling, quantitative analysis.
WIT Transactions on Engineering Sciences, Vol 72, 2011 WIT Press
www.witpress.com, ISSN 1743-3533 (on-line)
doi:10.2495/MC110041
42 Materials Characterisation V
1 Introduction
Nanotechnology has been considered one of the most promising branches on the
scientific development since the 60s [1] and its contributions to various fields
are notorious nowadays. The nano suffix and its effects became an intrinsic
subject of materials science and technology on the last few years as the unique
properties from nanometer scale are becoming increasingly relevant in those
fields.
The traditional materials engineering perspective [2], depicted in Figure 1,
can also be applied for the analysis of nanostructured systems. This approach
considers that the design of materials aiming an optimum performance requires
the complete understanding of the correlation among the processing routes
parameters, the materials structure and the resultant properties. However,
regarding nanostructured systems, this correlation may be stated as the
interrelationship among the synthesis parameters, the atomic scale morphology
and the energy configuration [3]. As the surface area to volume ratio is enhanced
for such systems, unique properties related to the surface energy distribution
become relevant. These properties are the source for the unique performance of
nanomaterials [4].
Figure 1:
Materials Characterisation V
43
ISSUE
Sample Preparation
Sample Stability
Energy Resolution
Signal to Noise Ratio
Sampling
Averaging
Data Analysis
EFFECT
modification on the sample original features
characteristics changes during analysis
spectroscopic signature may have overlaps
poor detection system sensitivity
lack of statistical representativeness
lack of features from individual particles
deconvolution, modeling and fitting
44 Materials Characterisation V
Transmission Electron Microscopy (TEM) based techniques stand out among
other characterization techniques for nanostructured systems due to the
possibility of a comprehensive and versatile approach including high resolution
imaging, diffraction and spectroscopy. This multipurpose characteristic allied
with simulation procedures provides complementary information that can be
used to improve the analysis precision.
Recent developments on the TEM instrumentation and analysis methods
allowed outstanding advances on nanoscience and nanotechnology [12]. The
spherical aberration (Cs) correction [13] can be considered among the most
influential improvements on both high resolution TEM (HRTEM) and high
resolution scanning TEM (HRSTEM) due to a number of factors. For HRTEM,
the Cs-correction of the objective lens resulted in the drastic improvement on
achievable spatial resolution, the possibility of larger gaps in the pole piece for in
situ experiments, and new imaging modes [14] due to fine tuning of the
aberration coefficients. For HRSTEM, the Cs-correction of the condenser lens
leads to an improvement on the achievable spatial resolution due to the probe
size reduction and the effective beam current increase.
In addition to the instrumental TEM improvements, the development of more
accurate models to describe the image formation and to support the TEM image
simulation [15] provides a more reliable interpretation of experimental data.
This work presents an overview on the HRTEM techniques state of art with
several examples which indicate the development of such characterization tools
towards the quantitative high resolution analysis of nanocrystalline materials.
Materials Characterisation V
45
Figure 2:
(a) SnO2 nanocrystal HRTEM image with an indexed FFT inset and
a HRTEM multislice simulation (red square). (b) oriented
attachment along the [110] is identified as the main growth
mechanism. [From ref. [20], copyright 2011 by RSC Publishing,
reprinted with permission of authors.]
46 Materials Characterisation V
Figure 3:
(a), (e) Trimetric and (b), (f) [111] zone-axis oriented views for the
Sb-doped nanocrystals geometric model. (c), (g) Original HRTEM
images and (d), (h) superimpositions with the simulated HRTEM
images. [Adapted from ref. [22], copyright 2009 by ACS
Publications, reprinted with permission of authors.]
Materials Characterisation V
47
Figure 4:
(a) Example of an original HRTEM image and (b) the output from
a FSR restoration from a CoSi2 sample. The insets exhibit a higher
magnification detail with the crystal structure superimposition.
[Adapted from ref. [29], reprinted with permission of authors.]
48 Materials Characterisation V
Figure 5:
Figure 6:
Materials Characterisation V
49
Figure 7:
Raw HRTEM images from the same CeO2 nanocrystal from a (a)
standard TEM microscope (1.25 information limit) and from a
(b) Cs-corrected microscope (0.8 information limit).
50 Materials Characterisation V
including the high resolution assessment of light elements atomic columns
information, such as occupancy next to crystalline defects [38] and sub-angstrom
displacement in magnetic domain walls [39].
Examples of NCSI HRTEM characterization for CeO2 sintered and
nanocrystalline samples are depicted on Figure 8. Although the contrast for
Oxygen columns is observed in both images, the direct analysis of a bulk-like
sample HRTEM image (Figure 8a) is unfeasible due to a combination of factors.
As the accurate thickness estimation is not achievable by HRTEM image
simulation for this specific combination of crystalline system and imaging
conditions, the intensity distribution cannot be directly associated to the Oxygen
columns scattering. A solution for this issue is the sample projected potential
assessment by the FSR application on NCSI HRTEM images (Figure 8b).
Figure 8:
5 Concluding remarks
State of art HRTEM techniques indicate that their use in conjunction with
theoretical procedures for HRTEM image simulation and data analysis represents
a vital tool for the quantitative evaluation of nanostructured systems. The
depicted examples illustrate how this technique can simplify, or even present
completely novel approaches, a number of materials characterization challenges
where a quantitative high resolution assessment is required.
The current development of HRTEM instrumentation and related
methodologies indicates that its importance on materials characterization will
grow on this decade, especially due to the novel possibilities achieved by
aberration correction implementation and by on-growing combined use with
HRSTEM imaging and spectroscopy techniques.
Materials Characterisation V
51
Acknowledgements
The authors acknowledge the financial support of the Brazilian research funding
agencies FAPESP and FINEP; the German research funding agency DAAD. The
authors would also like to thank the Ernst Ruska Centre staff at the
Forschungszentrum Jlich and the QFA staff at the Universitat Jaume I for the
fruitful discussions and scientific support.
References
[1]
[2]
[3]
[4]
[5]
[6]
[7]
[8]
[9]
[10]
[11]
[12]
[13]
[14]
Feynman, R. P.; Theres plenty of room at the bottom; Eng. Sci., 23(5),
22-36, 1960.
Callister, W. D.; Fundamentals of Materials Science and Engineering;
John Wiley & Sons: New York, pp. 2-5, 2001.
Burda, C.; Chen, X.; Narayanan, R.; El-Sayed, M. A.; Chemistry and
properties of nanocrystals of different shapes; Chem. Rev., 105(4),
1025-1102, 2005.
Kolmakov, A.; Moskovits, M.; Chemical Sensing and Catalysis by OneDimensional Metal-Oxide Nanostructures; Annu. Rev. Mater. Res.; 34,
151-180, 2004.
Shibata, N.; Pennycook, S. J.; Gosnell, T. R.; Painter, G. S.; Shelton, W.
A.; Becher, P. F.; Observation of rare-earth segregation in silicon nitride
ceramics at subnanometre dimensions; Nature; 428, 730-733, 2004.
Brune, H.; Giovannini, M.; Bromann, K.; Kern, K.; Self-organized growth
of nanostructure arrays on strain-relief patterns; Nature, 394, 451-453,
1998.
Yang, H. G.; Sun, C. H.; Qiao, S. Z.; Zou, J.; Liu, G.; Smith, S. C.;
Cheng, H. M.; Lu, G. Q.; Anatase TiO2 single crystals with a large
percentage of reactive facets, Nature, 453, 638-641, 2008.
Dalmaschio, C. J.; Ribeiro, C.; Leite, E. R.; Impact of the colloidal state
on the oriented attachment growth mechanism; Nanoscale, 2, 2336-2345,
2010.
Yu, W. W.; Wang, Y. A.; Peng, X.; Formation and Stability of CdTe
Nanocrystals; Chem. Mater.; 15, 4300-4308, 2003.
Barnard, A. S.; Modelling of Nanoparticles: Approaches to Morphology
and Evolution; Rep. Prog. Phys., 73, 1-52, 2010.
Brundle, C. R.; Evans, C. A.; Wilson, S.; Encyclopedia of Materials
Characterization, Butterworth-Heinemann, Reed Publishing USA, 1992.
Hawes, P.; Advances in Imaging and Electron Physics: Aberrationcorrected Electron Microscopy; Elsevier, 153, 2008.
Haider, M.; Uhlemann, S.; Schwan, E.; Rose, H.; Kabius, B.; Urban,
K.; Electron Microscopy Image Enhanced; Nature, 392, 768-769, 1998.
Jia, C. L.; Lentzen, M.; Urban, K.; High-resolution transmission electron
microscopy using negative spherical aberration; Microscopy and
Microanalysis, 10(2), 174-184, 2004.
52 Materials Characterisation V
[15]
[16]
[17]
[18]
[19]
[20]
[21]
[22]
[23]
[24]
[25]
[26]
[27]
[28]
Materials Characterisation V
[29]
[30]
[31]
[32]
[33]
[34]
[35]
[36]
[37]
[38]
[39]
53
Materials Characterisation V
55
Abstract
Owing to conversion of lime into calcium carbonate, lime is usually adopted for
conservative surface treatments. However, some critical aspects concerning the
treatments reduced penetration depth, the binder concentration and the
incomplete lime carbonation process still represent undesired limits and
hindrances. In order to improve lime treatments, Ca(OH)2 particles with
nanometric dimensions (nanolime) have recently been introduced in Cultural
Heritage conservation (frescoes, stuccoes, ..). The aim of the present work is to
characterize Ca(OH)2 nanoparticles synthesized by a chemical precipitation
process starting from two base supersaturated aqueous solutions of calcium
chloride and sodium hydroxide. After several washes, necessary to remove the
sodium chloride, the aqueous medium is partially substituted by 2-propanol to
improve the suspension stability; an alcoholic nanolime suspension is obtained,
characterised by a given concentration and a residual water content. In order to
identify the structure of the formed phases and the particles reactivity, the
obtained nanolime is characterised by X-ray diffraction (XRD) and profile
analysis; scanning and transmission electron microscopy (SEM-TEM) are
performed too. The results show hexagonally plated and regularly shaped
particles with side dimensions equal to or less than 300nm; moreover particles
have pure crystalline features and a high reactivity in terms of the carbonation
process. Finally, the Ca(OH)2 nanoparticles are applied on some natural
lithotypes; SEM analyses are performed to evaluate penetration depth and grain
adhesion of the nanolime treatments itself. From SEM micrographs a partial
filling of the lithotypes pores, located at a distance from the surface of more than
200m, is observed.
Keywords: calcium hydroxide, consolidation, lime, nanoparticles, protection.
WIT Transactions on Engineering Sciences, Vol 72, 2011 WIT Press
www.witpress.com, ISSN 1743-3533 (on-line)
doi:10.2495/MC110051
56 Materials Characterisation V
1 Introduction
Thanks to conversion of lime into calcium carbonate, lime water and milk are
usually adopted for conservative surfaces treatments. In particular, lime water
consolidation is generally obtained spraying the lime solution on the cleaned
surface. To reach a good penetration, the treatment is repeated several times until
the surface is able to absorb lime water [1]; some Authors indicate that could be
necessary to repeat the application for 30-40 times [2]. Lime milk is used on the
same basis as lime water [3]; nevertheless it involves greater amounts of lime
with the same water volume, reducing the water percentage brought to the stone.
However, some critical aspects concerning the treatments reduced penetration
depth, the binder concentration and the incomplete lime carbonation process still
represent undesired limits and hindrances. In order to improve lime treatments,
Ca(OH)2 particles with submicrometric dimensions (nanolime) are recently
introduced in Cultural Heritage conservation. Lime nanoparticles present the
following advantages in stone, mortar and plaster consolidation: the possibility to
penetrate deep into damaged zones (no limitations due to the particle size), high
reactivity and fast reactions (such as carbonation) in the treated zones, high
purity and defined composition [4]. Nanolimes are successfully employed on
mural paintings, stuccoes and frescoes [59]; refurbishments of architectonical
surfaces are considered too [1012]. Lime nanoparticles are typically produced
by a chemical precipitation process in supersaturated aqueous solutions of the
reactants (calcium chloride and sodium hydroxide). To improve nanolime
particles dispersion, the use of alcoholic solutions in place of aqueous ones is
adopted; in fact, when 2-propanol alcohol is used as a solvent, dispersions of
calcium hydroxide particles show a slower rate of agglomeration (and therefore,
slower sedimentation rates) in comparison to aqueous media. This reduces the
tendency for a white film to form on surfaces to be consolidated [13].
In this paper a nanolime suspension, characterised by a residual water content
in the precipitated phase, is synthesised. X-ray diffraction (XRD) measurements
and profile analysis are performed to characterise the nanolime samples and to
evaluate the carbonation efficiency (yield). To correlate the produced nanolime
to its properties, morphological characterisation is performed by scanning and
transmission electron microscopy (SEM and TEM).
Finally, the nanolime suspension is applied by brush on cleaned stone
surfaces of natural lithotypes; in particular, the suspension is used in a diluted
form (concentration of about 1mg/ml) in order to avoid the risk of leaving a
white film on treated materials surfaces [11].
Several lithotypes, largely employed in architectonic fields, are considered:
Travertine, Pietra di Lecce, Pietra Serena and Basalto.
A local calcareous lithotype, Poggio Picenze is investigated too. SEM
analyses allow one to investigate how the nanolime penetrates inside the stone
porosity, in terms of the reached depth and the grains adhesion of nanolime
treatments itself.
Materials Characterisation V
57
2 Experimental section
Calcium chloride dihydrate (CaCl22H2O), sodium hydroxide (NaOH) and 2propanol pro analysi products, supplied by Merck, are used without further
purification. Water is purified by a Millipore Organex system (R 18 M cm).
Two aqueous solutions of 400ml, containing 0,3mol/l of CaCl2 and 0,6mol/l
of NaOH respectively, are prepared. The NaOH solution (used as precipitator) is
added dropwise into the CaCl2 solution (speed 4ml/min, temperature of 90C).
After about 24 hours two distinct phases are observed: a limpid supernatant
solution and a white precipitated phase (NLW sample). In order to remove the
NaCl produced, several deionised water washings are performed. Subsequently,
in order to improve the suspension stability, the water content is partially
substituted by 2-propanol; the obtained suspension is characterised by a water/2propanol ratio (W/A) of 0,75 and a final concentration of about 10mg/ml (NLA
sample). SEM analyses (Philips XL30CP) are performed, depositing 0,2ml of the
suspension on the specimen; TEM investigations (Philips CM100) are carried
out dispersing 0,2 ml of the suspension in 50 ml of 2-propanol and depositing the
sample on the suitable grid.
As concerns XRD measurements, the sample is prepared maintaining the
nanolime preparation for 20 in ultrasonic bath (US) and then depositing 0,2ml
of the suspension on a silica sample holder; measures are performed on dry
sample, in laboratory conditions (T=20C, relative humidity RH=40%). Each
experimental diffraction spectrum is elaborated by a Profile Fit Software (Philips
PROFIT v.1.0) and each crystalline phase is attributed by JCPDS patterns; the
ratio between the CaCO3 peaks area and the spectrum total area is assumed as the
carbonation process efficiency (yield).
2.1 Results and discussions
TEM micrographs, obtained on NLA sample, are reported in Fig.1; in particular,
hexagonally and regularly shaped particles of side dimension ranging from 100
to about 300nm are observed. Moreover, in Fig.1b) it is possible to note that all
the particles appear overlapped in an ordered way and are so thin to be
transparent to the electron beam.
a)
b)
Figure 1:
58 Materials Characterisation V
In Fig. 2, SEM images on NLA sample are shown. The particles, characterised
by a prismatic features are recognizable to calcium carbonate of side dimension
less than 500nm (Fig.2a). In Fig.2b) a perfectly hexagonal and regular shaped
particle (marked as A) is observed, characterized by a side dimension less than
250nm.
a)
b)
SEM micrographs on NLA sample (scale bar 1m).
Figure 2:
XRD pattern on NLA sample shows the presence of Ca(OH)2 and CaCO3
phases (84-1276 and 85-1108 JCPDS patterns, respectively) with a
corresponding yield value of about 80% (Fig.3a). On the contrary, XRD results
obtained on NLW sample show as the corresponding yield decreases, reaching a
value of about 50% (Fig.3b). The NLA higher yield values can be attributed to
the presence of 2-propanol, that tends to disagglomerate the nanolime
particles; so, a greater specific surface of the Ca(OH)2 particles exposed to air,
can leads to a better carbonation process.
*
* CaCO3
Ca(OH)2
* CaCO3
Ca(OH)2
a)
b)
Figure 3:
Materials Characterisation V
59
a)
b)
Figure 4:
60 Materials Characterisation V
B
C
a)
b)
C
c)
d)
Figure 5:
a)
b)
Figure 6:
Considering the treated sample, the treatment covers homogeneously the internal
stone pores without filling them completely (Fig.7b-c). In particular, the
treatment is also well recognisable in the pore located at 800m from the surface,
underlining the penetration depth reached from the treatment itself (Fig.7c).
Materials Characterisation V
61
C
B
a)
b)
C
c)
Figure 7:
a)
b)
Figure 8:
62 Materials Characterisation V
SEM micrographs referred to treated stone, show as the nanolime treatment
tends to fill the internal cavity of the material, adhering to its grain, without
occlude the pores. In particular, the treatment, recognisable in a pore located at a
distance of 300m from the surface (Fig.9), is not found in more internal
cavities.
B
B
a)
b)
B
c)
Figure 9:
Materials Characterisation V
63
a)
b)
Figure 10:
B
a)
b)
C
c)
Figure 11:
64 Materials Characterisation V
a)
b)
Figure 12:
C
a)
b)
B
c)
d)
Figure 13:
SEM micrographs on the treated stone, show as the nanolime treatment tends
to fill the internal cavity of the material, adhering to its grain, without occluding
the porosimetric system. In particular, the treatment, is not found in pores more
internal than 700m.
Materials Characterisation V
65
4 Conclusions
Nowadays lime nanoparticles are synthesised and employed for protection and
superficial consolidation of several artworks such as stones, plasters, frescoes,
wall paintings and paper documents.
The obtained Ca(OH)2 nanoparticles are hexagonally plated and regularly
shaped, with side dimensions generally less than 300nm, as shown by TEM
observations. From XRD analyses, the carbonation process efficiency of the
aqueous nanolime is about 50%, while the use of 2-propanol improves it
reaching a yield value of about 80%. This result underlines the 2-propanol role in
disagglomerating the particles, leaving a higher specific surface exposed to air.
SEM analyses are performed to evaluate the interaction between the nanolime
treatments and the porosimetric system of some natural lithotypes. The images
show the partial filling of the pores in all the examined lithotypes, without
occluding them. On the contrary, considering all the lithotypes, the treatment
penetration is observed at a limit distance from the surface, ranging from 300m
(Poggio Picenze stone) to about 1mm (Travertine stone).
References
[1] Vinardi, M.G., Cully, M.H., Brunetto, A., La reversibilit nel restauro.
Riflessioni, esperienze, percorsi di ricerca, In: XI National Congress
Scienza e Beni Culturali: La Pulitura delle Superfici dellArchitettura,
Bressanone, XIX, pp. 399-405, 2003.
[2] Sten, P., Lime Water Consolidation, ICCROM, 53-61, 1981.
[3] Turriziani, R., I leganti ed il calcestruzzo, Edizioni Sistema, Roma,
pp. 70-82, 1972.
[4] Pianski, J., Brmmer, K., Ziegenbalg, G., Nano-particles for stone
conservationstate of the art, characteristics and recent developments, In:
Stonecore Recent progress in the consolidation of calcareous
materials, Litomyl, Czech Republic, Aprile 2010.
[5] Baglioni, P., Cesari, C., Dei, L., Giorgi, R., Grassi, R., Lorenzetti,
M., Mauro, M., Pinzani, D., Ruffo, P., Schonhaut, G., Stucco forte
veneziano con Tecniche della Tradizione: propriet chimico-fisiche,
degrado in ambiente lagunare e conservazione mediante nanofasi cristalline
di Ca(OH)2, In: National Congress Scienza e Beni Culturali: Lo Stucco.
Cultura, Tecnologia, Conoscenza, Bressanone, XVII, pp. 289-298, 2001.
[6] Ambrosi, M., Baglioni, P., David, P.R., Dei, L., Giorgi, R., Lalli,
C., Lanterna, G., Mairani, A., Matteini, M., Rizzi, M., Schonhaut,
G., Inorganic consolidants and protectives for architectonic surfaces:
experimental tests on Santa Prisca Church Apse in Rome, In: 2nd
International Congress on Science and Technology for the safeguard of
Cultural Heritage in the Mediterranean Basin, Vol. 2, pp. 873-877, 1992.
[7] Dei, L., Salvadori, B., Arlango, E., Pietropoli, F., Scardellato, C., Gli
affreschi del XIII e XIV secolo nella Cripta di San Zeno a Verona: la
sperimentazione della nanocalce dispersa in alcol iso-propilico durante
WIT Transactions on Engineering Sciences, Vol 72, 2011 WIT Press
www.witpress.com, ISSN 1743-3533 (on-line)
66 Materials Characterisation V
[8]
[9]
[10]
[11]
[12]
[13]
[14]
Materials Characterisation V
67
Abstract
Nanocarbon materials have numerous unique features high porosity, large
specific surface area, chemical inertness, radiation stability, etc. We applied
nanocarbon/nanodiamond and silicon carbidecomposites as matrix for optical
chemochips construction. Composite elements consist of porous nanocarbon
substrate with specific chromophores introduced to nanosize pores and siliconorganic coating. Similar multicomponent composition has a response to
radiation, in particular, -irradiation. By using diarylethenes as sensitive
chromophores, their electronic and/or luminescent spectra data may be applied
for doze power detection. Quantum chemistry methods, computer simulation
were considered for optimal design of nanocarbon/organic chromophore hybrid.
Experimental data and modeling have shown that diarylethenes are able to
change color under -irradiation in molecular crystal phase only. Weak
interactions (inter-, intramolecular and binding with hydrogen-containing walls
in pores) play key role under irradiation. Induced self-organization in limited
volume is considered. SWAXS and AFM data are discussed. Nanodiamond
composition possesses luminescent response to -, - and -irradiation. Siliconorganic polymer and SiC composition are neutral to irradiation, they are stable in
extreme conditions. Application of composite elements with optical sensing as
multifunctional chemochip fragment for atmospheric media monitoring is
discussed.
Keywords: nanocarbon, nanodiamond, silicon, composite, chemochip,
chromophore, radiation, computer simulation, optical response, nuclear waste.
68 Materials Characterisation V
1 Introduction
One of the most important directions of modern nanoscience and
nanotechnology is the so-called sensor technology combining the latest
achievements in highly sensitive intelligent materials, first of all, multifunctional
chemochips with immobilized sensing fragments with nanoscale function [14]
and different microdevice construction with high processing speed and
reliability. Of especial interest are the materials with high sensitivity and
adaptability [5]. In this context, great attention is being given to the selfassembling systems capable of hierarchical ordering from the nano- to
mesoscopic level [6], which may lead to elaboration of intelligent, in particular,
optically sensing, materials responsive to weak external factors such as
temperature, irradiation, pH, pollutants in the environment, etc. Typically, these
devices (chemochips) consist of solid inorganic support and sensing fragments
immobilized on it.
Nanocarbon materials are promising inorganic supports as they have a large
variety of unique and specific properties high porosity, large specific surface
area, chemical inertness, radiation stability and others [7]. To date, highly porous
(up to 40% vol.) composites can be prepared on the basis of nano-sized diamond.
Diarylethenes seem to be the most suitable objects to be used as sensing units.
These are photochromes changing from colourless to coloured under irradiation; normally, the colour disappears upon UV or visible-light irradiation
[8]. Diarylethenes have numerous advantages such as thermal stability, linear
dose dependence of the coloration, compared to other photochromic compounds.
On the other hand, diarylethenes exhibit their sensing properties in highly
ordered state (e.g. in single crystal). In this context, of great importance is
prediction of diarylethene molecules self-organization pattern.
Computer modeling and quantum chemistry calculations help greatly to
successfully solve applied science problems of predicting optimal synthetic
pathways, structure and properties of materials resulting in decrease in time and
costs of initial laboratory investigations. For integrated solution of the problem it
was necessary to predict diarylethenes self-organisation in a limited volume i.e.
to estimate the possibility for chromophores introduction into porous composite
and to evaluate the assembly parameters.
In this study, we applied experimental techniques together with a set of up-todate computing chemistry methods, Cambridge Structural Database and
specialized software. We obtained models of diarylethene molecules and their
assemblies; non-covalent interactions providing formation of 2D and 3D
supramolecular clusters and crystal structures were revealed.
Materials Characterisation V
69
F
F
F
F
F
H 3C
CH3
H3C
Open form
colourless
Figure 1:
F
F
F
F
UV or visible-light
irradiation
CH3
H3C
-irradiation
CH3
S H3C
CH3
Closed form
coloured
70 Materials Characterisation V
band is approximately 1520% from intensity of the luminescence peak of
starting diarylethene.
A preliminary study has shown rather high radiation stability of nanocarbon
composites , silicon-organic coating and chromophores.
Diarylethene molecules are conjugated aromatic systems capable not only of
hydrogen bonding but also of stacking interactions [12]. Pore size and shape also
influence self-organization of chromophores when filling the limited volume.
A molecular geometry modeling for diarylethene molecule and for its
molecular aggregates was performed by semi-empirical quantum chemistry
methods 1 by programme complex GAMESS [13]. The initial approximation
of the molecule (derived from x-ray diffraction data) was taken from Cambridge
Structural Database [14]. After optimizing the geometry parameters of
diarylethene molecule, we have estimated maximal distances between atoms
centres along the axes of three main inertias. These distances were ~9,9 (along
axis 1), ~11,5 (along axis 2), ~6,6 (along axis 3)
The optimization of geometry parameters of two diarylethene molecules
assembly was performed by non-empirical quantum chemistry method GAMESS
DFT 6-31G(d,p) [13]. These methods are also suitable for calculation of bond
critical points, in particular, bond critical points of non-covalent interactions
[15].
The size of the optimized model of the molecular aggregate including two
chromophore molecules (Fig. 2) have also been estimated by maximal distances
between atom centres along three axes. Two diarylethene molecules are arranged
in two parallel planes and are bonded by weak bonds. One can assume two
probable aggregate structures one is formed by four FH bonds and the
other by two SH bonds. Both aggregates dimensions are close to one
another these are 19* 11* 9 .
The proposed four chromophore molecules aggregate structure (Fig. 3) has
maximal distance between atom centres along one of the axes, equal to 29 .
The distance along two other axes is 18 and 9 .
As a result of the calculation performed, it was found that two diarylethene
molecules are also arranged in two parallel planes. In order to reveal weak noncovalent interactions in two diarylethene molecules assembly we have applied R.
Bader quantum topological analysis of atoms in molecules [15]. By programme
complex AIM-2000 bond critical points were calculated and a molecular graph
was drawn for two diarylethene molecules assembly [16]. We have calculated
bond critical points between carbon atoms, distances and angles of weak noncovalent interactions, local characteristics in critical points (Table 1). This
calculation has elucidated weak bonds between fluorine and hydrogen atoms,
these bonds are pairwise symmetrical. Moreover, weak bonds were found inside
the molecule, by bonds fluorine hydrogen an intramolecular cycle is formed.
Materials Characterisation V
71
(a)
(b)
Figure 2:
Table 1:
CP
number
1
2
3
4
5
6
7
8
9
2.54
2.53
2.53
2.54
2.60
2.60
2.78
2.78
4.20
Angle,
degrees
139.5
146.8
146.8
139.4
108.5
108.5
138.3
138.3
-
Contact
CH38F45
CH38F47
CH77 F8
CH77 F6
CH61F47
CH22 F8
CH22H70
CH38C23
C21C60
CP
number
55
48
52
35
36
24
43
47
33
2b
gb
vb
Econt
a.u.
a.u.
a.u.
a.u..
kcal/mole
0.00636
0.00645
0.00645
0.00635
0.00677
0.00677
0.00195
0.00195
0.00117
0.02822
0.02825
0.02825
0.02822
0.03319
0.03320
0.00619
0.00619
0.004150
0.005769
0.005794
0.005794
0.005768
0.006470
0.006471
0.001114
0.001114
0.0007669
0.004483
0.004525
0.004525
0.004482
0.004642
0.004642
0.0006811
0.0006811
0.0004963
1.41
1.42
1.42
1.41
1.46
1.46
0.21
0.21
0.16
72 Materials Characterisation V
(a)
(b)
Figure 3:
3 Conclusions
Experimental techniques (e. g. AFM, x-ray scattering and others) applied
together with a set of up-to-date computing chemistry methods aided greatly for
subsequent creation of a chemochip sensitive to -irradiation based on
WIT Transactions on Engineering Sciences, Vol 72, 2011 WIT Press
www.witpress.com, ISSN 1743-3533 (on-line)
Materials Characterisation V
73
Acknowledgements
The research is being supported by Ministry of Education and Science or
Russian Federation (Project 2.2.2.2.325) and ISTC (Project #3891).
References
[1] Al-Azzawi A. (ed). Photonics: Principles and practices, CRC Press: Boca
Raton, 2007.
[2] Davies A.G. and Thompson J.M.T. (eds). Advances in nanoengineering:
Electronics, materials and assembly, Imperial College press: London, 2007.
[3] Fryxell G.E. and Cao G. Z. (eds). Environment application of
nanomaterials Synthesis, sorbents and sensors, Imperial College Press:
London, 2007.
[4] Lu G.Q. and Zhao X.Z. (eds), Nanoporous materials: Science and
Engineering, Imperial College Press: London, 2004.
[5] Ong K. G., Yang X., Mukherjee N., Wang H., Surender S., Grimes C.A. A
wireless sensor network for long-term monitoring of aquatic environments:
Design and implementation, Sensor Letters, 2 (1), pp. 48-57, 2004.
[6] Grimes C.A, Dickey E.C., Pishko M. V., Encyclopedia of Sensors, ASP
Press: N.-Y., 2004.
[7] Seki Y., Impact of low activation materials on fusion reactor design,
J. Nucl. Mater, 258-263, pp. 1791-1797, 1998.
[8] Irie S., Irie M., Radiation-induced coloration of photochromic
dithienylethene derivatives in polymer matrices, Bull. Chem Soc. Jpn., 73,
pp. 2385-2388, 2000.
[9] Lisichkin G.V. (ed.) Chemistry of Grafted Surface Compounds, Fizmatlit:
Moscow, 2003.
[10] Gordeev S.K., Nanocarbon Materials, Nanotechnics (Rus.), 2005, pp. 3-11.
[11] Morimoto M., Kobatake S. Irie. M., Photochromism of diarylethenes in
nanolayers of a single crystal, Photochem. Photobiol. Sci., pp. 1088-1094,
2003.
WIT Transactions on Engineering Sciences, Vol 72, 2011 WIT Press
www.witpress.com, ISSN 1743-3533 (on-line)
74 Materials Characterisation V
[12] Yamada T., Kobatake S., Muto K., Irie M., X-ray Crystallographic Study
on Single-Crystalline Photochromism of Bis(2,5-dimethyl-3-thienyl)
perfluorocyclopentene, J. Am. Chem. Soc., 122, pp. 15891592, 2000.
[13] http://classic.chem.msu.su/gran/gamess/index.html.
[14] CSDB, http:// www.ccdc.cam.ac.uk
[15] Bader R., Atoms in Molecules: A Quantum Theory, Oxford University
Press, 1994.
[16] http://www.aim2000.de
[17] Espinosa E., Alkorta I., Rozas I., Elguero J., Molins, E., Topological
Analysis of the Electron Density Distribution in Perturbed Systems. I.
Effect of Charge on the Bond Properties of Hydrogen Fluoride, Chem.
Phys. Letts., 336, pp. 457-464, 2001,.
Section 3
Corrosion problems
Materials Characterisation V
77
Abstract
Fretting-corrosion is one of the main concerns in the application of hip
prosthesis. This type of information is very important in the stage of orthopedic
implants design, with the purpose of minimizing the amount of tissue exposed to
corrosion products which are released during the permanence of the prosthesis in
the patient. The residual corrosion products of stainless steels are associated to
the occurrence of several adverse reactions in the human body. The knowledge
about these corrosion products is extremely important in the phase of the project
of hip prostheses, aiming at the minimization of the amount of exposure of the
organic tissues to corrosion products released during the permanence of the
prosthesis inside the patient.
In the present work the mechanical stability and fretting corrosion resistance
of modular hip prosthesis, which was fabricated with ASTM F 138 austenitic
stainless steel, were evaluated according to the criteria of ASTM F 1875
standard, method I, which prescribes long term test, with the purpose of
determining the amount of damage through the quantification of the corrosion
products and debris which resulted from the fretting corrosion conditions. The
mechanical tests were performed in a servohydraulic mechanical testing machine
and the modular interfaces were exposed to an electrolytic 0.9% NaCl in distilled
water solution and subjected to a minimum load of 230 N and a maximum load
WIT Transactions on Engineering Sciences, Vol 72, 2011 WIT Press
www.witpress.com, ISSN 1743-3533 (on-line)
doi:10.2495/MC110071
78 Materials Characterisation V
of 2.3 kN, frequency of 5 Hz for 10 million cycles, according to ASTM F 1875
standard, thus simulating real conditions of use. Load ratio (R = 10) was
determined in ASTM F1440 standard. A significant amount of fretting-corrosion
products were observed: 0.22 g resulted from five specimens after 107 million
cycles, and the diffraction X-Rays tests showed the presence of the crystalline
phases Fe2O3, Fe3O4 and Cr2O6.
It can be concluded that the mass loss in the head-cone connection allowed
the entrance of the physiological saline solution in the inner region of the head,
increasing the predicted micro movement in the headcone interface, resulting in
an accelerated process of fretting-corrosion, and consequently the liberation of
debris and corrosion products that could lead to adverse biological reactions.
Keywords: fretting corrosion, hip prosthesis, modular components, mechanical
properties.
1 Introduction
The increasing life expectancy in the last decades and trauma risk due to the
more frequent exposure of individuals to different types of accidents and current
insecure life conditions has led to an increased search for orthopedic implants
with high performance, capable of resisting to even more severe loads for longer
times. For this reason, metallic materials, which were developed for applications
in orthopedic implants, must present some specific properties, such as
biocompatibility, strength and resistance to degradation (by wear or corrosion).
In spite of the increasing use of titanium and its alloys for this application in
the last decades, in Brazil ASTM F138 austenitic stainless steel [1, 2] is still very
used in the fabrication of modular components for total hip arthroplasty due to its
lower cost, together with good mechanical properties and satisfactory corrosion
resistance, since the main customer is Health Ministry, which supplies the
demands of public hospitals in Brazil.
The use of modular components in the total hip arthroplasty presents several
advantages, such as a great variety of stem geometry, which allows a better
choice of modular component for the surgeon, according to the patient disease,
mainly in the case of revision surgeries, thus minimizing the risk of inadequate
procedures.
However, the modular interfaces are subjected to micro movements, which
can result in fretting and corrosion, leading to the release of debris, which can
cause adverse reactions and accelerated wear in the articulation interface.
Fretting can be defined as a wear phenomenon which occurs when two solids
in contact are subjected to tangential oscillatory movement with small
displacement amplitude. It can be described more accurately as a movement
whose amplitude is smaller than the contact extent. The presence of a corrosive
environment contributes to accelerate the wear process and this condition is
defined as fretting-corrosion. The occurrence of corrosion results in
degradation processes which reduce the structural integrity of the implant and
release products whose reaction can be harmful in contact with organic tissue.
Materials Characterisation V
79
When the prostheses are fabricated with ASTM F138 stainless steel, the
generation of a passive barrier in the surface (passive layer) is the main obstacle
to corrosion process. This film, which is composed by metallic oxides, formed
spontaneously on the metal surface, and whose thickness can be increased with
surface treatments, avoids the migration of metallic ions from the metal to the
solution and the migration of anions from the solution to the metal through the
metal-solution interface (physiological environment). For being effective barriers
these films must be compact and cover completely the metallic surface, they
must have an atomic structure which limits the migration of metallic ions and/or
electrons through the metallic oxide/solution interface, and they must be capable
of remain on the surface of these alloys even when subjected to mechanical and
abrasive stresses in the clinical use of prostheses.
The most frequently mentioned mechanisms of failure in these components
are aseptic loosening, periprosthetic osteolysis and metallosis and are directly
related to the debris released by the prosthesis components, mainly the non
cemented prostheses [3].
According to the literature [4], the residual products of corrosion in stainless
steel based implants are associated to tissue necrosis, inflammation, allergenic
reactions and even cancer.
The aim of the present work is an evaluation of fretting corrosion in modular
prostheses fabricated with ASTM F 138 stainless steel, in order to minimize the
adverse reactions during the permanence of the prosthesis in the patient.
Head
Figure 1:
Cone
Stem
Head and stem of the hip prosthesis fabricated with ASTM F 138
stainless steel.
80 Materials Characterisation V
The fretting corrosion tests on the cone-head interface were performed
according to method I of ASTM F 1875 standard, which prescribes a 0.9% NaCl
physiological saline solution, thus simulating in vivo conditions, since that when
implanted in patients these components are in contact with body fluids. With the
purpose of analyzing both fatigue resistance of hip stems and mechanical
stability of modular components the experimental procedures described in
ASTM F 1440 [5] and ISO 7206-4 [6] were also adopted.
These tests were performed in two universal machines for mechanical testing
(Instron model 8872) with a 25 kN load cell with the following characteristics,
described in ASTM F 1875 standard [7]:
-
Frequency: 5 Hz;
Computerized operation;
Materials Characterisation V
Figure 2:
81
X-ray diffraction analysis was carried out in the following conditions: anode
material: Cu K, angle between 10 and 100, step size 0.05, scan step time: 1 s,
receiving slit size: 0.03 mm, current: 40mA, voltage: 40 kV.
Figure 3:
Detailed view inside the heads: (a) before and (b) after the test.
Magnification: 10X.
82 Materials Characterisation V
Head-cone interface
Figure 4:
Figure 5:
Figure 6:
Materials Characterisation V
83
Figure 8: Detail.
Until 1 million cycles neither corrosion products nor particulate debris could
be observed, but after this period the degradation of the modular component was
increasing and localized on the head-cone, as can be seen in detail in Fig. 5.
After 10 million cycles the corrosion products and particulate debris were
quantified, resulting in an average value of 0.22 g. In the literature there is only
one citation of fretting-corrosion products quantification, but in their experiment
the cycle number was varied, thus hindering any type of comparison with the
results presented in this work.
For better comprehension of fretting corrosion mechanisms in hip modular
components the variation of load and number of cycles is necessary. The
characterization of the corrosion products and particulate residues by X-ray
analysis is also very important, in order to identify the phases and type of
metallic particles which were released, with the aim of minimizing the incidence
of aseptic loosening, periprosthetic osteolysis and metallosis in the surgeries of
total hip arthroplasty. In the present work a complete characterization of the
fretting corrosion products was not concluded, but it can be observed that they
presented a predominantly ellipsoidal shape with size distribution approximately
uniform, whose larger diameter was around 500 m.
WIT Transactions on Engineering Sciences, Vol 72, 2011 WIT Press
www.witpress.com, ISSN 1743-3533 (on-line)
84 Materials Characterisation V
Anyway, the X-ray diffraction analysis performed in the corrosion products
showed the presence of the following crystalline phases: Fe2O3, Fe3O4 and
incipient Cr3O8, compatible with the ASTM F138 stainless steel.
4 Conclusion
The evaluation of fretting corrosion on the head-cone connection interfaces of
five modular hip prosthesis fabricated with stainless steel led to the following
results:
A significant grade of localized corrosion in the inner region of the head
and in the head-cone interface was observed, originated from the
fretting corrosion and fretting during 10 million cycles, resulting in
particulate debris and corrosion products.
In order to achieve a better comprehension of the fretting corrosion
mechanisms, more tests are necessary, including changes in the applied
load and number of cycles.
At first glance, fretting corrosion does not affect the mechanical
stability of the modular components, at least until 107 cycles.
References
[1] Cavalcanti, E.H. de S., Souza, S.M.C. de, Ferreira, C. de A., Campos, M.M.,
Abud, I. de C. and Palmeira, L., (2002), Avaliao da Resistncia
Corroso de Prtese Total de Quadril de Ao Inoxidvel Austentico
Removida de Paciente, 22 CONBRASCORR Congresso Brasileiro de
Corroso, Salvador Bahia.
[2] Giordani, E.J., Ferreira, I. and Balancin, O., (2007), Propriedades
mecnicas e de corroso de dois aos inoxidveis austenticos utilizados na
fabricao de implantes ortopdicos, REM: R. Esc. Minas, Ouro Preto,
60(1): 55-62.
[3] Crestani, M.V., Boschim, L.C. and Schwartsmann, C.R., (2004), Metalose
Simulando Tumor Abdominal , Revista Brasileira de Ortopedia, Nov/Dez.
2004.
[4] Merritt, K.; Brown, S.A.; (1985), Biological Effects of Corrosion Products
from Metals, Corrosion and Degradation of Implant Materials: Second
Symposium, ASTM STP 859, A.C. Fraker and C.D. Griffin, Eds., American
Society for Testing and Materials, Philadelphia, 1985, pp. 195-207.
[5] ASTM F 1440 92 (Reapproved 2008), Standard Practice for Cyclic
Fatigue Testing of Metallic Stemmed Hip Arthroplasty Femoral Components
Without Torsion.
[6] ISO 7206-4-2010 - Implants for surgery - Partial and total hip joint
prostheses -- Part 4: Determination of endurance properties and performance
of stemmed femoral components.
[7] ASTM F 1875 98 (Reapproved 2004), Standard Practice for Fretting
Corrosion Testing of Modular Implant Interfaces: Hip Femoral Head-Bore
and Cone Taper Interface.
WIT Transactions on Engineering Sciences, Vol 72, 2011 WIT Press
www.witpress.com, ISSN 1743-3533 (on-line)
Materials Characterisation V
85
Abstract
Based on the idea that hydrophobic (low or non-wettable) surfaces can decrease
the contact area between a corrosive solution and a surface, thereby potentially
rendering the material more corrosion resistant, the effect of surface patterning
on the corrosion behaviour of nickel was investigated. The surface patterning
consisted of an array of holes of various diameters (D) and inter-hole spacings
(L) that were produced by a laser ablation process. The corrosion behaviour of
the patterned surfaces was studied using a potentiodynamic polarization method
in a 0.5M H2SO4 electrolyte and compared with that of a polished reference
sample. Following the potentiodynamic polarization corrosion test cycle, the
corroded surfaces were examined using scanning electron microscopy (SEM) for
morphological features and white light interferometry (WLI) to determine the
surface roughness. The changes in surface morphology were related to the
corrosion behaviour. A relationship was found between D, L, and the corrosion
current density (Icorr), whereby the higher the (D/L )2 ratio, the higher the Icorr
value. The corrosion potential (Ecorr) of all surface patterned samples was lower
(less noble) than that of the reference sample in all tests.
Keywords: corrosion resistance, surface patterning, hydrophobicity, laser
ablation.
1 Introduction
Wettability of solid surfaces is an important property and depends both on the
surface chemistry and on the surface topology. Recently, hydrophobic and
superhydrophobic surfaces (water contact angles larger than 150) have received
a lot of attention, due to their important applications ranging from self-cleaning
WIT Transactions on Engineering Sciences, Vol 72, 2011 WIT Press
www.witpress.com, ISSN 1743-3533 (on-line)
doi:10.2495/MC110081
86 Materials Characterisation V
materials and in-flight icephobic treatment [1] to microfluidic devices. The
archetype superhydrophobic surface is that of the lotus leaf, on which a water
droplet apparently forms a sphere, reducing the area of contact. This behavior,
known as the lotus-leaf-effect, is found to be a result of the hierarchically
patterned structure. Effectively the fluid cannot penetrate the air gaps between
the patterned pillars on the surface, and therefore it forms a heterogeneously
wetted contact surface of alternating liquid/solid contacts and air pockets. As a
result the overall solid liquid contact area is significantly reduced [25].
In the present work we employ surface patterning to create heterogeneous
wetting on metallic surfaces. We show that heterogamous wetting in metallic
surfaces leads to reduction in corrosion rates and increase in corrosion resistance.
2 Experimental procedures
2.1 Specimen preparation
Pure nickel (99.7 Wt.%) was selected as a model metal. Samples of 1.5 cm 1.5
cm size were polished to a standard finish with a roughness that did not exceed
50 nm. Laser ablation with a single pulse copper bromide (CuBr) metal vapour
laser was used to create special surface textures. During laser ablation, nitrogen
(N2) was blown to protect the surfaces from oxidation and clean any debris. The
pulse duration was selected at 30 ns. Different hole/pattern sizes were achieved
with different laser output power within the range of 20-80 W. Figure 1 shows a
schematic of the applied pattern. The diameter of the holes and the inter-hole
spacing were varied to obtain different textures. Table 1 presents the selected
hole size and the inter-hole spacing. The distances between the holes are labeled
L1, L2 and L3, where as the D1, D2, D3 etc are the hole diameters.
Figure 1:
Hole diameter
(D) m
5
10
20
Inter-hole
spacing
L2, m
10
20
30
Inter-hole
spacing
L3, m
20
30
40
Materials Characterisation V
87
For easy identification of the samples the following sample labeling system
has been adopted: DxLy, where x is the diameter of the hole in m and the y is
the inter-hole spacing in m.
2.2 Corrosion tests
The corrosion resistance of the samples was determined using a potentiodynamic
polarization corrosion test, wherein the samples were immersed into a 0.5M
H2SO4 solution at room temperature (24 C). A conventional three-electrode
system was used in which a standard calomel electrode (SCE) served as the
reference electrode and a platinum electrode as a cathode electrode. In the tests,
the applied potentials were in the range of -0.7 to 1.5 V (with respect to SCE)
with a scan rate of 1.0 mV/s. Corrosion rates were calculated in terms of the
corrosion current density, Icorr, by using linear polarization resistance curves
(LPR). The relationship between Icorr and the polarization resistance, Rp, is
obtained from the StearnGeary equation [5]:
Rp
where a and
a c
E
(1)
E
is polarization
iapp
resistance. After the corrosion test the sample surfaces were examined using
JEOL 5800 scanning electron microscopy (SEM), and energy-dispersive X-ray
spectrometry (EDS).
88 Materials Characterisation V
trends in Figure 2 suggest that a decrease of the hole size (D) m towards nanoscale size could decrease the corrosion current density, Icorr, even further.
Figure 3 presents the Ecorr values for samples with pattern density of (D/L)2=1
and (D/L)2=0.25. The corrosion potential (Ecorr) of all surface patterned samples
was lower (less noble) than that of the reference sample (1/D=0.0 m-1) in all
tests.
a)
b)
Figure 2:
The Icorr values versus inverse hole diameters (1/D) for all samples
with (D/L)2=1 and (D/L)2=0.25 after the corrosion test.
a)
b)
Figure 3:
The Ecorr values versus inverse hole diameters (1/D) for all samples
with (D/L)2=1 and (D/L)2=0.25 after the corrosion test.
Materials Characterisation V
a)
89
b)
Figure 4:
a)
b)
Figure 5:
a)
b)
Figure 6:
90 Materials Characterisation V
Sample D20L40 exhibited the lowest amount of patterning degradation.
Comparing the original sample and after the corrosion test there was no obvious
change of the surface pattern [Figures 6(a) and 6(b)]. The measured Icorr for this
sample was the lowest of all samples including reference and patterned. No
surface damaged was detected also in samples D20L20 and D20L40. Clearly the
observed surface damage in all samples was in a good agreement with the
predicted trends by the measured Icorr values.
Sample REF
a)
Sample REF
b)
Figure 7:
Materials Characterisation V
91
Ni/O ration remained at 49/5, indicating that passive oxide layer was not formed.
Similar EDS spectra were obtained for all corrosion damaged samples. For
instance the EDS spectra of sample D5L20 on the surface [Fig. 8 (a)] and in the
hole [Fig. 8 (b)] after corrosion testing do not show any significant difference
and the Ni/O ratio is the same as for the polished REF sample. However, the
EDS spectra for the pattered samples with the improved corrosion properties
(D20L20, D20L30, and D20L40) exhibit different behaviour. The Ni/O ratios for
sample D20L40 before and after the corrosion test are shown in [Figs. 9 (a),(b)
and Figs. 10(a),(b)]. The concentration of the O on the surface before and after
the corrosion test slightly changed, which is consistent with the other corrosion
damaged samples.
However, the bottom of the patterned holes shows significant increase in O
concentration (Figs. 8(b) and 10(b)). This clearly indicates that Ni oxides were
formed but were not dissolved by the electrolyte. A possible explanation could
be that the fluid actually did not reach the bottom of the patterned hole. In other
words, in the samples with better corrosion resistance, the contact between the
electrolyte and the metal surface is heterogeneous wetting alternating
solid/liquid zones and air pockets.
Sample D5L20-surface
a)
Sample D5L20-hole
b)
Figure 8:
92 Materials Characterisation V
Sample D20L40-surface
a)
Sample D20L40-hole
b)
Figure 9:
a)
Sample D20L40-hole
b)
Figure 10:
Materials Characterisation V
93
4 Conclusions
Surface patterns with different pattern density were successfully created on Ni by
laser ablation. The patterns had a distinct effect on the corrosion properties of the
metallic surfaces. A specific group of patterns led to significant decrease of the
corrosion current density and the corresponding corrosion rate. For this group of
patterns it has been shown that a possible reason for the decrease in corrosion
rate is the detected heterogeneous wetting on the patterned surface.
References
[1] K. Varanasi, T. Deng, J. David Smith, M. Hsu, and N. Bhate, Frost
formation and ice adhesion on superhydrophobic surfaces Appl. Phys. Lett.
Vol. 97, 2010, No 234102.
[2] M. Shafiei and A. T. Alpas, Nanocrystalline nickel films with lotus leaf
texture for superhydrophobic and low friction surfaces, Applied Surface
Science Vol. 256, 2009, pp. 710719.
[3] W. Barthlott and C. Neinhuis, Purity of the sacred lotus, or escape from
contamination in biological surfaces, Planta, Vol. 202, 1997, pp. 1-8.
[4] S.J. Lee, C.H. Huang, J.J. Lai and Y.P. Chen, Corrosion-resistant
component for PEM fuel cells, Journal of Power Sources, Vol. 131, 2004,
pp. 162-168.
[5] T. Liu, Y. Yin and L. Dong, New application of the underwater superhydrophobic surface in the corrosion protection, Advanced Materials
Research, Vols. 79-82, 2009, pp. 1115-1118.
[6] M. G. Fontana and N. D. Greene, Corrosion engineering, 2nd edition,
McGraw Hill, 1978.
Materials Characterisation V
95
Abstract
Large military vehicles within museum collections are stored in two distinct
environments, controlled and uncontrolled, with an intermittent transitional
mode where vehicles travel between the two. Variable environmental conditions
combined with operational factors pose significant risks to the reliability,
durability and longevity of these vehicles. Although there are methods for
retarding or decelerating aspects of failure, to maintain the integrity and
originality of these vehicles as artefacts a sustainable methodology for
conserving these vehicles should be developed.
Corrosion is one of the significant contributors to the structural damage and
material aging of historical military vehicles; therefore an experimental study
was conducted to understand the prevailing mechanisms of failures due to
corrosion with various types occurring in these vehicles identified. This paper
presents various modes of corrosion in historic vehicles while X-ray
Fluorescence and ultrasonic scanning corrosion mapping techniques characterise
corrosion inhibiting materials and subsequent material loss. Understanding
material profiles and their link to environmental exposure during use and non use
of these vehicles will lead to a sustainable methodology for their conservation.
Keywords: sustainable, military, vehicle, corrosion, materials, conservation.
96 Materials Characterisation V
1 Introduction
The structural and functional deterioration of the large metal structures due to
their environmental exposure and the operating conditions are responsible for
their structural aging [1]. Aging mechanisms in large metal structures such as
bridges, aircrafts, vehicles etc is recognised as a growing problem [2]. Aging
mechanisms such as corrosion, surface and sub surface cracks, wear due to
interacting surfaces and undesired stresses are leading to mechanical failures,
higher maintenance costs and compromises the structural integrity. Corrosion
has been identified as one of the major contributor to the aging mechanisms; it
affects all metal structures by deteriorating the materials properties [3].
This research focuses upon the corrosion damage within the large historic
vehicles in the museum environment. Corrosion is indiscriminate in affecting
metal structures and museum artefacts are no exception. The military vehicles
from World War I and II were designed and manufactured according to the
automotive technologies in place at those times [4] with only a limited lifeexpectancy, however their participation in the war and the subsequent historic
significance means it is important to devise sustainable methodologies for longterm preservation. Continuous corrosion problems within large historic military
vehicles in the museum environment are leading to the concerns of their
longevity, in particular where failures in protective measures are also identified
[5]. For the purpose of sustainable methods it is important to understand the
various corrosion modes, material characterisation and the material loss profile
due to corrosion.
For this research military vehicles were inspected for visual signs of corrosion
and novel methods such as X-ray fluorescence for elemental identification and
ultrasonic scanning for material loss were conducted. Visual inspection
identified different modes of corrosion depending on their exposure to extreme
environments during the war, operating conditions at the time of their service
life, types of materials used and their current working or non-working
environments.
X-Ray fluorescence method found Fe, the main constituent; other elements
such as W, Mo and Mn were also identified. Ultrasonic methods found severe
material loss due to corrosion in the selected samples.
2 Experimental methodologies
The following methods are used in the research. Only the most common and
reoccurring types of corrosion which were found on a larger scale in the vehicles
are presented here.
2.1 Visual inspection
Visual inspection was completed on more than 20 large military vehicles from
the World War I, II and post war era. The investigation found one or more forms
of corrosion in the same vehicles. Some widespread and most occurring modes
of corrosion are described from four military tanks Table 1.
WIT Transactions on Engineering Sciences, Vol 72, 2011 WIT Press
www.witpress.com, ISSN 1743-3533 (on-line)
Materials Characterisation V
Table 1:
97
No
Vehicle
Environment
1
2
3
4
5
Mark II
Sherman*
King Tiger
Centurion
Scorpion
Controlled
Uncontrolled
Controlled
Uncontrolled
Uncontrolled
Manufacture
Country of
Origin
Era
1917
United Kingdom WW 1
1940-1941
United States
WW 2
1943
Germany
WW 2
1945
United Kingdom Post war
1973
United Kingdom Post War
3 Results
3.1 Uniform corrosion
With uniform corrosion a direct chemical attack affects the surface of the metal
and then spreads out evenly across the entire surface, results are not usually fatal
and it is not classified as one of the severe forms [9]. Figure 1 of the King
Tigers flywheel shows uniform corrosion spreading across the surface. Uniform
corrosion was also found in Tog II and FT 17 on considerably larger surface
areas. Uniform corrosion, if left to continue, could lead to general thinning of the
surface and may transform in to another form of corrosion [10].
3.2 Fretting corrosion
Fretting corrosion occurs at the interface of the interacting surfaces which are
subject to relative motions. The King Tiger planetary gears (Figure 2) are a
classic example of this form. Interacting mechanical wear lead to the removal of
material particles, followed by the oxidation of the material debris and the newly
exposed surface [11]. The oxidised debris would act as an abrasive media
leading to further wear and fatigue failures could occur [9, 12].
WIT Transactions on Engineering Sciences, Vol 72, 2011 WIT Press
www.witpress.com, ISSN 1743-3533 (on-line)
98 Materials Characterisation V
Figure 1:
Figure 2:
Materials Characterisation V
Figure 3:
Figure 4:
99
Figure 5:
Figure 6:
Pitting in centurion.
Materials Characterisation V
Figure 7:
Table 2:
101
Surface: Corroded
MQ for points: A= 5.70 B= 4.50 C=
4.60
06 American
06 American
Material Standard
06 American
Constituents %
A- Reading 254
B-Reading 255
C-Reading 256
4.89
0.46
0.19
0.05
0.11
5.92
0.54
0.19
0.05
0.08
5.38
0.85
0.2
0.05
0.09
1
2
3
4
5
Si
P
Ti
V
Cr
Mn
0.86
1.11
0.78
7
8
9
10
Fe
Co
Ni
Cu
92.68
0.21
0.06
0.05
91.48
0.22
0.08
0.03
92.05
0.12
0.1
0.04
11
12
Mo
W
0.05
0.33
0.06
0.2
0.06
0.12
Figure 8:
Figure 9:
Figure 10 shows the end views and the c-scan of the sample with
implementation of colour coding Red for 0.00 mm and Blue for 10.00 mm. Two
cursors Red and Blue show the X-axis, Y-axis location and the corresponding
depths of the particular points on the sample.
A user defined area is taken for finding the maximum and minimum
thicknesses starting at 2.00 mm on the X-axis, 2.00 mm on the Y-axis in
Figure 11 and finishing at 172.00 mm on the x-axis and 66.00 mm at the y-axis
in Figure 12 respectively to avoid any distortion in signals at the edges. The scan
detected maximum remaining thickness of 9.80 mm at horizontal position of
150.00 mm and vertical position of 26.00 mm with a drastic loss in material to a
minimum of 4.60 mm at horizontal position of 28.00 mm and vertical position of
8.00 mm. The maximum thickness of 9.80 mm is recorded at two different
points. The second lowest depth is 4.70 mm and is detected at seven different
WIT Transactions on Engineering Sciences, Vol 72, 2011 WIT Press
www.witpress.com, ISSN 1743-3533 (on-line)
Materials Characterisation V
Figure 10:
Figure 11:
103
positions; third lowest of 4.80 mm at six various points and the remaining
thickness of 4.90 mm are recorded at four different points in the samples. The
minimum remaining thickness of 4.60 mm is recorded at three different points.
4 Discussion
The modes of corrosion identified above can be retarded by using the protective
measures such as coatings and paint for uniform corrosion, application of
lubrication and exclusion of air in the fretting corrosion, alleviating the stress in
stress corrosion cracking, using same metal types in galvanic or applying
insulation between two different metals.
WIT Transactions on Engineering Sciences, Vol 72, 2011 WIT Press
www.witpress.com, ISSN 1743-3533 (on-line)
Figure 12:
The alloying elements such as Mn, Cr, V W and Mo identified though XRF
are used as hardening agents to improve hardness, ductility and tensile strength
[19]. Also Mo, Mn and P are used for the rust and corrosion prevention. Mo is
used in high strength alloys and super alloys for strength and wear
resistance [20].
Sherman M4A1 was designed by Lima Locomotive Works in the United
States. At the time of manufacture the armoured thickness range was kept
between maximum of 62.00 mm and a minimum of 12.00 mm [21, 22]. However
it is not practical to follow any possible modification in the armour thickness and
because of this reason the highest thickness recorded in the scan is considered
the maximum thickness. During the service life, exposure to extreme
environments has led to the deterioration and material loss. The ultrasonic scan
of has shown a remarkable material loss of 5.20 mm between the maximum
remaining thickness of 9.80 mm and the minimum remaining thickness of 4.60
mm. Currently in the museum collection, the Sherman travels between two
controlled and uncontrolled environments. Varying environments and the
likelihood of failing protective measures, temperature fluctuations, humidity
ratio variations, rain water and the extent of chemical reaction etc will influence
the rate of corrosion and in the controlled environment corrosion residues from
may result in indoor atmospheric corrosion [23, 24].
5 Conclusion
The research work uses a conventional method of analysing corrosion through
surveying vehicles in the museum environment which is followed by novel
methods of XRF for the elemental identification and Ultrasonic scanning to
measure and map corrosion. Research work will be conducted for corrosion
WIT Transactions on Engineering Sciences, Vol 72, 2011 WIT Press
www.witpress.com, ISSN 1743-3533 (on-line)
Materials Characterisation V
105
modelling and for the life expectancy of large historic vehicles based upon the
results from this research work in order to maximise the longevity, durability and
reliability.
Acknowledgements
The authors would like to thank Mark Nel, Mark Clark in AGR, Jon Previtt
West-Dean College and Mike Hayton in The Tank Museum Bovington for their
in kind support during this research work.
References
[1] M. Colavita, "Service Life and Aging of Military Equipment," Corrosion:
Environments and Industries (ASM Handbook), vol. 13C, pp. 220-228,
2006.
[2] A. S. Vinod, "Corrosion in the Military," presented at the Corrosion:
Environments and Industries (ASM International), 2006.
[3] Z. Ahmad, "Basic Concepts in Corrosion," in Principles of Corrosion
Engineering and Corrosion Control, Oxford: Butterworth-Heinemann,
2006, pp. 9-56.
[4] I. C. Handsy and J. Repp, "Ground Vehicle Corrosion," Corrosion:
Environments and Industries (ASM Handbook), vol. 13C, pp. 148-150,
2006.
[5] M. T. Gudze and R. E. Melchers, "Operational based corrosion analysis in
naval ships," Corrosion Science, vol. 50, pp. 3296-3307, 2008.
[6] B. Beckhoff, Handbook of practical X-ray fluorescence analysis. Berlin;
London: Springer, 2006.
[7] R. v. Grieken and A. A. Markowicz, Handbook of X-ray spectrometry, 2nd
ed., rev. and expanded. ed. New York, N.Y.: M. Dekker, 2002.
[8] J. C. Drury, Ultrasonic flaw detection for technicians, 3rd ed. [Swansea]:
Silverwing Ltd., 2004.
[9] R. Baboian, Corrosion tests and standards : application and interpretation.
Philadelphia, Pa.: ASTM, 1995.
[10] H. H. Uhlig and R. W. Revie, Corrosion and corrosion control : an
introduction to corrosion science and engineering, 4th ed. / R. Winston
Revie. ed. Hoboken, N.J.: Wiley-Interscience; Chichester: John Wiley
[distributor], 2008.
[11] R. C. Bill, "The role of oxidation in the fretting wear process," 1981.
[12] A. Neyman and O. Olszewski, "Research on fretting wear dependence of
hardness ratio and friction coefficient of fretted couple," Wear, vol. 162164, pp. 939-943, 1993.
[13] R. P. M. Procter, et al., "Stress-corrosion cracking of C---Mn steels in
methanol-ammonia environments--I. Effects of environmental and
mechanical variables," Corrosion Science, vol. 33, pp. 1009-1031, 1992.
[14] R. C. Newman, "Stress Corrosion Cracking," in Shreir's Corrosion, J. A. R.
Tony, Ed., Oxford: Elsevier, 2010, pp. 864-901.
WIT Transactions on Engineering Sciences, Vol 72, 2011 WIT Press
www.witpress.com, ISSN 1743-3533 (on-line)
Section 4
Computational models
and experiments
Materials Characterisation V
109
Abstract
A Multi-Factor-Interaction-Model (MFIM) is briefly described to represent
complex point material behavior in a single equation. The model is of product
form in order to represent coupled interactions and to be computationally
effective. The model describes a continuum or surface in space that represents
the complex material behavior in terms of the various factors that affect a
specified material behavior. The material specified behavior is inclusive of all
material properties, mechanical, thermal, physical and effects thereon, such as
temperature, time, cyclic loadings, etc. Sample case results simulated by using
MFIM are compared with test data to illustrate its versatility and its relevance to
reality. These results show that the MFIM can accurately predict metal matrix
composite fatigue data and mechanical properties of a steel alloy. Helpful
guidelines for its effective use are also included.
Keywords: material properties, high temperature, nonlinearities.
1 Introduction
The simulation of complex material behavior resulting from the interaction of
several factors (such as temperature, nonlinear material due to high stress, time
dependence, fatigue, etc), has been mainly performed by factor-specific
representations. For example, entire text books are devoted to plasticity, creep,
fatigue and high strain rate to mention only a few. Investigators have derived
equations that describe material behavior for each factor-specific effect.
Suppose we visualize that the material behavior is a continuum represented by
some surface. Then, we can think of some representation which describes that
surface which is inclusive of all participating factors that affect material behavior
either singly or interactively in various combinations. To that end, research has
WIT Transactions on Engineering Sciences, Vol 72, 2011 WIT Press
www.witpress.com, ISSN 1743-3533 (on-line)
doi:10.2495/MC110101
B
(2)
Ai 1
B0
Here B represents a specific cause factor for behavior (for example,
temperature), and Bo is the corresponding final value. This concept is
schematically represented in Fig. 1, the development of the multi-factorinteraction model (MFIM) to represent complex material behavior by a single
equation. The objective of the present paper is to briefly describe MFIM and
present results obtained there from to illustrate its uniqueness and its versatility.
Values for Bo and mi for specific behavior are selected either from known
behavior or more likely from a best judgment in conjunction with consultations
with seasoned professionals for that behavior. By representing the MBS with the
MFIM of product form (Eq. (1)), we gain another distinct advantage. The
behavior factors, B, can also be represented by another level of MFIM or
progressive substructuring of equation (1). The progressive substructuring leads
to a multitier representation of the MBS that permits intrinsic lower tier
behaviors to influence more than one factor at the next higher tier. In other
words, the observed specific behavior (Bi may depend on another set of lower
WIT Transactions on Engineering Sciences, Vol 72, 2011 WIT Press
www.witpress.com, ISSN 1743-3533 (on-line)
Materials Characterisation V
111
tier elemental behaviors). Further, the behavior factors in this lower set of
specific behaviors may depend on yet another next lower tier of elemental
behaviors. That is, there are usually sets and subsets of specific behaviors that
hierarchically influence the higher level behaviors. When this is done, N can be
limited to 6, (for example), but the number of factors influencing material
behavior at the next lower tier will increase exponentially as Nj where j is the
number of 6-factor tiers. For example, when j = 3, N = 216, and so forth. This
representation is natural for multiparallel processing computers where the tiers
are programmed with different granularities. Obviously, then, the motivation for
selecting such a form is for computational and programming effectiveness.
Another reason for selecting an MFIM of product form is that the effect of each
factor can be evaluated separately. The interpretation of Bo is that it represents a
scale, whereas mi represents a shape or path. For example, (1 B/Bo)mi where 1
> B/Bo and + < mi < , covers the whole space as is illustrated in Fig. 2.
Figure 1:
The inclusiveness of this particular form, combined with its simplicity, makes
it very attractive for a computational simulation. An expanded form of MFIM is
shown in Eq. (3) below:
m
Mp Tgw T
Mpo Tgw To
E
...1 e
E
ef
1
Sf
Ce
1
Cef
t
1
Sf tf
...
Cc
1
Ccf
M NM
1
Sf Nf M
T NT
1
Sf Nf T
1
f
...
(3)
Figure 2:
Materials Characterisation V
113
since the exponents will be the coefficients of linear algebraic equations. (5) To
simulate block fatigue evaluate the degradation up to the end of the first block
then use the degraded properties for the second block and repeat for additional
blocks degradation after second input to third etc. (6) Unloading generally
requires different exponent then loading. Select the unloading exponent to
simulate the expected loading-unloading by stress cycle. (7) Intermittent time
effects are handled the same way as block fatigue. (8) Combined fatigue (thermo
mechanical) is simulated by using the corresponding cycles and stresses in the
respective factors. (9) Frequency effects are simulated in combination with
cyclic loads. Stress value is not needed here since it is included in the cyclic
effects factor. (10) The static stress and the temperature factors when present
are used in combination with all active factors for that simulation effect. (11) To
simulate fabrication process use the temperature factor, the stress factor due to
pressure, a flow model and update the geometry through the finite element model
of the component. The user will invent other ways to use MFIM that the author
has not even thought of yet.
The most important point to remember is that MFIM is to be used
incrementally where the current values are updated from global thermo
structural, etc. analyses. This may require a two level iteration: (a) local to
achieve local equilibrium. During the iteration, the factors are changed until
local equilibrium is reached; and (b) global to achieve global equilibrium with
respect to boundary loads and supports conditions and internal stress field or
energy input in the increment equals the energy added to the structure. Another
important point to note is that the exponent for specific factor in MFIM
represents a general trend and not the entire precise path from its reference value
to its final value. Also, note that the MFIM is used for all the properties that is
the same factors but with different exponent. Results from sample cases for
MFIM application are described for: metal matrix composites (MMC), and a
nickel base alloy. As mentioned previously, these results are presented to
illustrate a few of the generic features of the MFIM applicability to convey to the
reader what has be done and infer therefrom what can be done. The specific
how is described in the references.
Figure 3:
Table 1:
Materials Characterisation V
115
Figure 4:
Figure 5:
Figure 6:
Figure 7:
Figure 8:
The time factor was activated for the fabrication process and the tensile load
to fracture. Additional simulation results are described in [3, 4].
Materials Characterisation V
117
Figure 9:
Figure 10: Calibration of nickel base alloy IN-100 specific heat as a function of
temperature.
WIT Transactions on Engineering Sciences, Vol 72, 2011 WIT Press
www.witpress.com, ISSN 1743-3533 (on-line)
6 Concluding remarks
The salient concluding remarks from a description of the multi-factor interaction
model (MFIM) to predict complex material behavior by using a single equation
and select applications are: (1) MFIM is a material point simulator. It is unique
and inclusive for both composites and metals. (2) It is used incrementally in
combination with analyses that provide local stress information. (3) It usually
requires alternative solutions consisting of both local and global convergence. (4)
Applications for meta matrix composites under different loading conditions
predicted results that were in very good agreement with test data. (5) Familiarity
and confidence and extensive usage will be gained by continuing usage
application to complex solutions of material behavior and to simulate difficult
and even untried problems.
References
[1] Chamis, C.C., Lark, R.F. and Sinclair, J.H.: Integrated theory for Predicting
the Hydrothermal Mechanical Behavior of Composite Structural
Components. ASTM pp. 160-192.
[2] Chamis, C.C. and Hopkins, D.A.: Thermo Viscoplastic Nonlinear
Constitutive Relationships Structural Analysis of High Temperature Metal
Matrix Component. NASA TM 87291, 1985
[3] Chamis, C.C., Murthy, P.L.N. and Hopkins, D.A.: Computational
Simulation of High Temperature Metal Matrix Composites Cyclic Behavior.
ASTM, STP 1080, 1990, pp 5669.
[4] Tong, M.T., Singhal, S.N., Chamis, C.C. and Murthy, P.L.N.: Simulation of
Fatigue Behavior of High Temperature Metal Matrix Composites. ASTM
Reprint from Standard Technical Publication, 1253, 1996, pp. 540-551.
[5] Boyce, L. and Chamis, C.C.: Probabilistic Constitutive Relationships for
Cyclic Material Strength Models. AIAA.ASME/ASCE/AHS 29th Structures,
Structural Dynamics and Materials Conference. Part 3, AIAA, 1988,
pp. 1299-1306.
[6] Progressive Fracture Structural Analysis of National Wind Tunnel
Structures by L. Minnetyan, NASA CR 198485, May 1996.
Materials Characterisation V
119
Abstract
Functionally Graded Materials (FGMs) possess properties that vary gradually as
a function of spatial coordinates. They are different from conventional composite
materials in that they have no distinct interfaces at which their material
properties change abruptly. These FGMs are suitable for various applications,
such as aerospace, nuclear fusion, biomaterial electronics, etc. In practice,
applications of analytical solutions are limited. However, the analytical solutions
are very important as standards for evaluating numerical simulation results and
they are also important to mathematical understanding. Little research on the
analytical solutions of two-dimensional elastostatic problems has been reported.
Furthermore, few analytical solutions using Airy stress functions have been
published. The purpose of this paper is to propose an analytical method for the
two-dimensional elastostatic problems of FGMs using the Airy stress function.
In the present investigation, FGMs in which the properties of the materials vary
exponentially in one direction are examined. A few numerical examples are
presented and the validity of the method is shown by comparisons with the
results of past studies.
Keywords: analytical solution, functionally graded material, two-dimensional
problem, Airy stress function.
1 Introduction
The Functionally Graded Materials (FGMs) possess properties that vary
gradually as a function of spatial coordinates. They are different from
conventional composite materials in that they have no distinct interfaces at which
WIT Transactions on Engineering Sciences, Vol 72, 2011 WIT Press
www.witpress.com, ISSN 1743-3533 (on-line)
doi:10.2495/MC110111
xz zz
0
x
z
where xx , zz , zx are stress components and zx xz .
xx zx
0
z
x
(1)
Materials Characterisation V
121
p z ( x) p 0
h
l
z
Figure 1:
xx
where
u
x
xx , zz , zx are
zz
w
z
zx
u w
z x
(2)
2 xx 2 zz 2 zx
0
zx
z 2
x 2
(3)
xx s11 xx s13 zz
where
zz s13 xx s33 zz
zx s44 zx
(4)
s11 , s13 , s33 , s 44 are elastic moduli. The material properties of FGMs
s ij s F ( z ) s e h
(5)
0
ij
0
ij
xx
2
2
z
zz
x 2
zx
zx
(6)
(
2
)
2
13
44
33
11
h z 3
z 4
x 2 z 2
x 4
2
2
3
2
2
0
0
0
s
s
(2 s130 s 44
)
0
11
13
h x 2 z
h 2 z 2
h 2 x 2
s110
(7)
The above equation is the same form as the governing equation of the plates
bending with variable rigidity, and it is possible to adopt the same method for the
solution [10].
3 Method of solution
In Eqn. (7), we introduce the following Airy stress function ( x,z ) .
( x, z ) Ai ( z ) f i ( x)
(8)
i 1
Ai ( z ) f si ( x )
i 1
where
A ( z ) sin x
i 1, 3, 5
(9)
i
, f si ( x) sin i x . And the stress components are as follows
l
d 2 Ai ( z )
xx
f si ( x)
dz 2
i 1, 3, 5
zz
i 1, 3, 5
zx
i 1, 3, 5
2
i
Ai ( z ) f si ( x )
dAi
f ci ( x)
dz
(10)
Materials Characterisation V
123
obtained easily.
3
2
2
d 4 Ai
2 d Ai
0 d Ai
0
0
0
s
s
s
s
(
2
}
11
11
13
44
i
h dz 3
dz 4
h2
dz 2
2
2 dAi
0
0
(2 s130 s 44
) i
( s33
i 2 s130 2 ) i 2 Ai 0
h dz
h
s110
(11)
Ai ( z ) C i ,1e i ,1 C i , 2 e i , 2 C i ,3 e i , 3 C i , 4 e i , 4
r z
where the
r z
r z
r z
(12)
2
2
1
3
0
) i2 ri , j
ri , j 2 0 (2 s130 s 44
h
s11
h
1 0 2
2
0
0 (2 s130 s 44
) i2 ri , j 0 ( s 33
i s130 2 ) i2 0
h
s11
s11
h
ri , j 2
4
(13)
p z ( x) p0 , the
pz
4 p0 1
1
f si ( x)
sin i x
i 1,3,5 i
i 1,3,5 i
4 p0
(14)
zz 0
zx 0
at z h 2
(15)
Substituting Eqn. (10) and Eqn. (14) into Eqn. (15), the following four equations
for determining the arbitrary constants C i , j are derived.
4 p0
h
Ai ( )
2
i 2 i
h
Ai ( ) 0
2
dAi h
( ) 0
dz
2
dAi h
( )0
dz 2
(16)
p z ( x) p0
h
l
z
Figure 2:
A2 ( z ) f 2 ( x) A2 ( z ) x 2
where
(17)
xx
zz
zx
d 2 A2 2
2 d 2 A2
f
(
x
)
x
2
z 2
dz 2
dz 2
d 2 f 2 ( x)
2
A
(
z
)
A2 ( z ) 2
2
x 2
dx 2
dA ( z ) df 2 ( x)
dA ( z )
2
2
2
2x
zx
dz
dx
dz
(18)
Substituting Eqn. (17) into Eqn. (7), the following ordinary differential
equation is derived again.
3
2
2
d 4 A2
0 2 d A2
0
0
0
2 d A2
s x
2s11 x
{2(2s13 s44 ) s11 2 x } 2
h
dz 4
dz 3
h
dz (19)
2
0 dA2
2(2s130 s44
)
2s130 2 A2 0
h dz
h
0 2
11
Materials Characterisation V
125
0 dA2
2(2s130 s44
)
2s130 2 A2 0
h dz
h
2
0
11 0
The solution of the ordinary differential equation is given in the same form as
Eqn. (12). The boundary conditions on the top z h 2 and on the bottom
h
A2 ( ) 2 p 0
2
h
A2 ( ) 0
2
dA2 h
( ) 0
dz
2
dA2 h
( )0
dz 2
(21)
4 Numerical examples
4.1 Analysis of simply supported FGM beam under sinusoidal pressure
Let us consider a simply supported FGM beam subjected to a sinusoidal pressure
as shown in Fig. 3. For convenience, the results are indicated in the coordinate
system of Fig. 3 [6]. The Youngs modulus is assumed to be of the form
E E0 e z h and the constitutive equation is the same as that of Ref. [7].
h
l
z
Figure 3:
Two types of the material properties are considered, and the ratios of Youngs
moduli of the top surface and the bottom surface are Eh E0 =10 and
Eh E0 =0.1. In the case of the former, the graded index in Eqn. (5)
corresponds to 2.30 , and in the case of the latter, it corresponds to
=-2.30 . Further, in the case of the former, the load is applied on the softer
WIT Transactions on Engineering Sciences, Vol 72, 2011 WIT Press
www.witpress.com, ISSN 1743-3533 (on-line)
Galerkin method
xx(x,z)/xx(x,h)
2.0
0.0
0.0
0.2
0.4
0.6
0.8
1.0
-2.0
-4.0
-6.0
z/h
Figure 4:
Axial stress
xx ( x, z ) / xx ( x, h)
beam at x l 2 .
For the graded index
component
xx
3 ,
coordinate z and the vertical axis is the stress values. In this figure, Ref. [9]
means the results of Ref. [9]. The present results again are in good agreement
with those of Ref. [9]. It was also confirmed that the distributions of the stress
components zz and zx are also in good agreement.
WIT Transactions on Engineering Sciences, Vol 72, 2011 WIT Press
www.witpress.com, ISSN 1743-3533 (on-line)
Materials Characterisation V
127
Galerkin method
xz/xz(average)
2.0
1.5
1.0
0.5
0.0
-0.5 0
0.2
0.4
0.6
0.8
z/h
Figure 5:
Shear stress
xz / xz (average)
beam at x l 2 .
Table 1:
Materials properties.
Material constants
Values [1/Pa]
5.41 10 12
1.51 10 12
0
11
0
13
0
s33
9.52 10 11
0
s44
1.37 10 10
graded index =+3, -3
xx(l, z)
Present (=+3)
Present (=-3)
Ref.[9] (=+3)
Ref.[9] (=-3)
250.0
200.0
150.0
100.0
50.0
0.0
-50.0 -0.5 -0.4 -0.3 -0.2 -0.1 0.0 0.1
-100.0
-150.0
-200.0
-250.0
z/h
Figure 6:
0.2
0.3
0.4
0.5
xx (l , z ) ( 3 ).
xx
0.0
-10.0 -0.5 -0.4 -0.3 -0.2 -0.1
0.0
0.1
0.2
0.3
0.4
0.5
-20.0
-30.0
-40.0
-50.0
-60.0
-70.0
z/h
Figure 7:
5.0
4.5
4.0
zx
3.5
3.0
2.5
2.0
1.5
1.0
0.5
0.0
-0.5
-0.3
-0.1
0.1
0.3
z/h
Figure 8:
0.5
Materials Characterisation V
xx
xx
xx
129
zx
is different from
that for homogeneous beams. The point where the maximum stress appears
moves toward the softer surface. It may be possible to control distribution of
stresses by adjustment of materials properties.
5 Concluding remarks
In this paper, one analytical solution method for two-dimensional elastostatic
problems in FGMs in which the properties of the materials vary exponentially in
one direction is proposed. The method is based on the idea that the governing
equation of this problem is of the same form as the governing equation for the
plate bending with variable rigidity. The analysis method uses the Airy stress
function. From analysis of a few numerical examples, most of the results
correlate well with other solutions and the validity of the method is shown. The
distributions of displacement components are not stated, however we can obtain
them by integrating the strains and displacements relationships.
Future work will be focused on development of an analytical method of
solution for analyzing problems of FGMs having an arbitrary variation of their
material properties expressed by functions other than exponential.
References
[1] Suresh, S., and Moretensen, A., Fundamentals of Functionally Graded
Materials, IOM Communications Ltd, London, 1998.
[2] Chakraborty A., Gopalakrishnan S. and Reddy JN., A new beam finite
element for the analysis of functionally graded materials, International
Journal of Mechanical Science, Vol.45, pp.519-539, 2003.
[3] Trung-Kien Nguyen, Karam Sab and Guy Bonnet, First-order shear
deformation plate models for functionally graded materials, Composite
Structures, 83, pp.25-36, 2008.
[4] H.J. Xinag and J. Yang, Free and forced vibration of a laminated FGM
Timoshenko beam of variable thickness under heat conduction,
Composites: Part B Vol.39, pp.292-303, 2008.
[5] Sankar BV., An elasticity solution for functionally graded beams,
Composite Science Technology, 61, pp.689-696, 2001.
[6] Zhu H. and Sankar BV., A combined Fourier series-Galerkin method for
the analysis of functionally graded beams, Journal of Applied Mechanics,
Vol.71, pp.421-424, 2004.
Materials Characterisation V
131
Abstract
The moment-curvature (M-) diagrams define the maximum capacity of
structural elements that are primarily subjected to bending moments and
therefore their failure mode is flexural. The moment-curvature diagrams are also
used to assess the ductility of structural elements and are therefore very
important for the determination of the amount of plastic energy a structural
element can absorb. One way of improving the reinforced concrete (RC)
members is the confinement of concrete with FRP composites. The work
presented in this paper includes the development of moment curvature diagrams
for RC members using a fiber model. The RC members were designed as
compression controlled members meaning that their failure initiates in the
concrete prior to yielding of the steel tension reinforcement. These types of
members have limited ductility. The introduction of a specific amount of Carbon
Fibre Reinforced Polymer (CFRP) composite grid as confining reinforcement
improves the ductility of the RC members by as much as 30%. The results from
the fiber model are compared to available results from an experimental program
conducted to evaluate the experimental improvement of the ductility of
compression controlled members. The experimental and analytical results are a
good match indicating that the fiber model is accurate and can be used to develop
the moment-curvature diagrams of RC members confined with a CFRP
composite grid.
Keywords: CFRP grid, concrete confinement, moment curvature, ductility.
1 Introduction
Compression controlled steel reinforced members, fail through crushing of
concrete and therefore lack the ductility that is associated with yielding of the
WIT Transactions on Engineering Sciences, Vol 72, 2011 WIT Press
www.witpress.com, ISSN 1743-3533 (on-line)
doi:10.2495/MC110121
2 Experimental investigation
The unique application of the CFRP composite grid as confining reinforcement
was evaluated with the construction of two beams. Beam 1 had no CFRP
composite grid tubes and served as the control beam and beam 2 had two CFRP
WIT Transactions on Engineering Sciences, Vol 72, 2011 WIT Press
www.witpress.com, ISSN 1743-3533 (on-line)
Materials Characterisation V
133
composite grid tubes placed in the compression zone. Details of the dimensions,
amount and location of steel reinforcement and CFRP composite grid tubes are
provided in Figures 1 and 2. The CFRP composite grid was formed into a
circular tube using a 152 mm diameter plastic pipe and was held in place by a
thin rope wrapped around the tube along its length.
5 in (127 mm)
3 in (76 mm)
2.75 in
(70 mm)
18 in
(457 mm)
6.65 ft (2.03 m)
2.7 ft (0.81 m)
6.65 ft (2.03 m)
16 ft (4.87 m)
1 in (25 mm)
1 in (25 mm)
#3 (#10) Stirrups
18 in (457 mm)
8 #10 (#32) Bars
2.75 in (70 mm)
2 in (51 mm)
3 in (76 mm)
12 in (305 mm)
Figure 1:
5 in (127 mm)
2.75 in
(70 mm)
18 in
(457 mm)
6.65 ft (2.03 m)
2.7 ft (0.81 m)
6.65 ft (2.03 m)
16 ft (4.87 m)
0.75 in (19 mm)
1 in (25 mm)
1 in (25 mm)
6 in
(152 mm)
#3 (#10) Stirrups
18 in (457 mm)
8 #10 (#32) Bars
2.75 in (70 mm)
2 in (51 mm)
3 in (76 mm)
12 in (305 mm)
Figure 2:
Spreader Beam
D15
D14
D13
D12
D11
D10
D9
D8
D7
D6
D5
D4
D3
D2
D1
18 in
(457 mm)
North
Support
South
Support
5.5 ft (1.7 m)
LVDT
3 ft (0.91 m)
Tilt Meter
Figure 3:
5.5 ft (1.7 m)
60 mm Foil Gauge
Test set-up.
Materials Characterisation V
135
the strains in that area close to the neutral axis were small and therefore the
effect on the strength was minor. The unconfined region was divided into one
rectangular layer while the confined region was divided into eight rectangular
layers. For the control beam, the compression zone was treated as an unconfined
concrete region and was divided into eight layers.
1 in. (25 mm)
c - 25 mm
(c - 1 in.)
Grid Tube
N. A.
Unconfined concrete
Confined concrete
Figure 4:
Some of the assumptions employed in this fiber model include perfect bond
between concrete and the reinforcing bars (strain compatibility), plain sections
remain plain (linear strain distribution), the area below the neutral axis was
considered cracked and was ignored in the force and moment calculation and
monotonic loading and deformation of the section.
The compressive force in each layer was calculated as the product of the area
of the layer and the average stress. The average stress was determined based on
the strain in the layer and the stress-strain curves of concrete. The Hognestad
parabola was used to calculate the stress-strain curve of unconfined concrete.
The modified Hognestad was used to calculate the stress-strain curve of confined
concrete [10]. The parameters of the modified Hognestad were determined based
on the available data from confined concrete cylinder tests [9].
The tensile force was calculated as the product of the area of steel
reinforcement and the steel stress. Steel stress was determined based on the
average strain of the two steel layers and by assuming elastic perfectly plastic
stress-strain curve.
The moment in the cross-section for the applied curvature distribution was
determined by summing moments about the neutral axis.
The moment curvature diagram was calculated by applying increasingly
larger top concrete compressive strains (c) to the cross-section and varying the
location of the neutral axis (c) until equilibrium of the axial forces (zero) in the
WIT Transactions on Engineering Sciences, Vol 72, 2011 WIT Press
www.witpress.com, ISSN 1743-3533 (on-line)
39
500
678
400
542
300
407
200
271
Experimental
Fiber Model
100
136
0
0
0.2
0.4
0.6
0.8
Curvature x 1000 (rads/in)
Figure 5:
Moment (kip-ft)
Moment (kN-m)
39
500
678
400
542
300
407
200
271
Expiremental
Fiber Model
100
136
0.2
0.4
0.6
0.8
Curvature x 1000 (rads/in)
Figure 6:
Moment (kN-m)
Moment (kip-ft)
Materials Characterisation V
137
tube beam compared to the control beam, although relatively small, was evident
when the experimental M- curves for both beams were compared (see Figure
7). The area under the load-displacement curve represents the amount of energy
a structure can absorb before failure. The amount of energy is also a good
indicator of ductility. Ductile structures usually can absorb higher amounts of
energy compared to non-ductile (brittle) structures.
39
500
678
400
542
300
407
200
271
Control
Grid
100
Moment (kN-m)
Moment (kip-ft)
136
0.2
0.4
0.6
0.8
Curvature x 1000 (rads/in)
Figure 7:
Experimental M- curves.
Figure 8:
d y or d c
du
The curvature ductility () was approximately 1.5 and 2 for the control and
tube beam respectively. The schematic for determining u and c for the control
beam is shown in Figure 9. The same approach was used in determining the
ductility factor for the tube beam. The curvature ductility factor of the tube beam
was approximately 33% higher than the curvature ductility factor of the control
beam.
WIT Transactions on Engineering Sciences, Vol 72, 2011 WIT Press
www.witpress.com, ISSN 1743-3533 (on-line)
Materials Characterisation V
39
678
400
Mu
Mi
542
300
0.8 Mu
0.75 Mi
407
200
271
100
136
c
0
0
Figure 9:
0.2
0.4
0.6
0.8
Curvature x 1000 (rads/in)
Moment (kN-m)
500
Moment (kip-ft)
139
0
1
4 Conclusions
In this paper the development of moment-curvature diagrams for compression
controlled RC sections confined with CFRP composite grids was presented.
These diagrams were compared with diagrams developed from experimental data
for the same sections. Based on the results from this comparison the following
conclusions can be drawn:
1. The results from the two beams indicate that the CFRP grid provided
confinement to the concrete. Confinement was achieved using a series
of small diameter CFRP grid tubes rather than wrapping the entire
cross-section.
2. The curvature ductility factor of the beam utilizing the CFRP grid tubes
was improved by 33% compared to the control beam with no grid tubes.
The improvement was significant, considering the small amount of
CFRP that was used to confine the compression zone.
3. The energy dissipation of the tube beam was 37% higher compared to
the controlled beam due to the effect of concrete confinement. This
increase in the energy dissipation of the element also represents an
improvement in the ductility due to concrete confinement from the
CFRP grid tubes.
WIT Transactions on Engineering Sciences, Vol 72, 2011 WIT Press
www.witpress.com, ISSN 1743-3533 (on-line)
5.
The results from the fiber model are fairly accurate compared to the
experimental data and therefore the model is considered valid.
However, there is a divergence after the peak load and further
refinement of the model is required.
The number of test specimens is limited and therefore the more testing
is required. Due to the small number of test specimens the conclusions
are drawn with great caution.
References
[1] Rooney, J., and Taylor, S., 2004, Flexural Behaviour of Steel and GFRP
Reinforced Concrete Beams, Composites: Part B, Vol. 27, No 3-4,
pp. 38-40.
[2] Toutanji, H., and Deng Y., 2003, Deflection and Crack Width Prediction
of Concrete Beams Reinforced with Glass FRP Rods, Construction and
Building Materials, Vol. 17, No 1, pp. 69-74.
[3] Grace, N. F., Soliman, A. K., Sayed, G. A., and Saleh, K. R., 1998,
Behavior and Ductility of Simple and Continuous FRP Reinforced
Beams, Journal of Composites for Construction, Vol. 2, No 4,
pp. 186-194.
[4] Rahman, A. H., Kingsley, C. Y., and Kobayashi, K., 2000, Service and
Ultimate Load Behavior of Bridge Deck Reinforced with Carbon FRP
Grid, Journal of Composites for Construction, Vol. 4, No 1, pp. 16-23.
[5] Yost, J. R., Goodspeed, C. H., and Schmeckpeper, E. R., 2001, Flexural
Performance of Concrete Beams Reinforced with FRP Grids, Journal of
Composites for Construction, Vol. 5, No 1, pp. 18-25.
[6] Harries, K. A., and Gassman, S. L., 2003, Load Tests of Reinforced
Concrete Catch Basing Knockout Panels, Department of Civil and
Environmental Engineering, University of South Carolina, Report No
ST03-01, p. 21.
[7] Shao, Y., Johnson, C., and Mirmiran, A., 2003, Control of Plastic
Shrinkage Cracking of Concrete with TechFab Carbon FRP Grids,
Department of Civil, Construction, and Environmental Engineering, North
Carolina State University, Report to Tech-Fab Inc, p.
[8] Hamilton, H. R, Cook, R. A., and Alfonzo, L., 2006, Crack Control in
Toppings for Precast Flat Slab Bridge Deck Construction, Department of
Civil and Coastal Engineering, University of Florida, Engineering and
Industrial Station, Final Report Submitted to FDOT, UF Project No
00030907, p. 120.
[9] Michael, A. P., Hamilton, H. R. III, and Ansley, M. H. (2005). Concrete
confinement using carbon fiber reinforced polymer grid, American
Concrete Institute, SP230-56, Vol. 230, pp. 991-1010.
[10] Park, R., and Paulay, T., 1975 (a), Chapter 6: Ultimate Deformation and
Ductility of Members with Flexure, Reinforced Concrete Structures, John
Wiley & Sons, New York, NY, pp. 195-269.
Materials Characterisation V
141
[11] Priestley, M. J. N., and Park, R., 1987, Strength and Ductility of Concrete
Bridge Columns under Seismic Loading, ACI Structural Journal, Vol. 84,
No 1, pp. 61-76.
[12] Zahn, F. A., Park, R., and Priestley, M. J. N., 1990, Flexural Strength and
Ductility of Circular Hollow Reinforced Concrete Columns without
Confinement on Inside Face, ACI Structural Journal, Vol. 87, No 2,
pp. 156-166.
[13] Sheikh, S. A., and Khoury, S. S., 1993, Confined Concrete Columns with
Stubs, ACI Structural Journal, Vol. 90, No 4, pp. 414-431.
[14] Yeh, Y. K., Mo, Y. L., and Yang, C. Y., 2002, Seismic Performance of
Rectangular Hollow Bridge Columns, Journal of Structural Engineering,
Vol. 128, No 1, pp. 60-68.
[15] Park, R., and Paulay, T., 1975 (b), Chapter 11: Strength and Ductility of
Frames, Reinforced Concrete Structures, John Wiley & Sons, New York,
NY, pp. 496-609.
Materials Characterisation V
143
Abstract
Hygric and thermal properties of several types of materials which are used in
reconstructions of historical stone masonry on Czech territory are experimentally
measured at first. The obtained data are then subjected to a homogenization
procedure, and the effective heat and moisture transport and storage parameters
of the typical fragment of masonry are calculated. For the homogenization, a
mixing model originally derived for applications in dielectric studies was
employed taking into account the theoretical bounds of the investigated material
parameters. The obtained data can find use in damage assessment of historical
masonry using methods of computational simulations and analysis.
Keywords: hygric properties, thermal properties, historical stone masonry,
argillite, sandstone, lime-based plaster, homogenization techniques.
1 Introduction
The hygrothermal behaviour of climatically exposed components and structures
of historical buildings is related to the hygric, thermal, mechanical and other
physical and chemical properties of inbuilt materials. Their knowledge
represents necessary information for the understanding of the building
performance as well as the first step in avoiding damage, degradation or the
undue heat loss from the constructions.
The damage assessment of historical masonry and other structures due to the
negative effects of moisture and temperature can be done effectively by means of
mathematical and computational modelling. In this way, the time development of
WIT Transactions on Engineering Sciences, Vol 72, 2011 WIT Press
www.witpress.com, ISSN 1743-3533 (on-line)
doi:10.2495/MC110131
Materials Characterisation V
145
2 Studied materials
Two types of stones that were in the past often used in Czech region as bearing
masonry materials are analysed together with lime-based mortar with pozzolana
admixture.
Many historical buildings in the Czech Republic were built using similar
kinds of sandstone. Siliceous raw-grained sandstone was usually used for
historical architectural constructions (walls, portals, window frames) for its
strength. Ornamental parts of the architecture (gothic flowers, romantic shells)
and sculptures (from the Romanesque period up to now) were made of finegrained calcite-argillaceous sandstone. In this work, sandstone from Men-lzn
quarry, Czech Republic, is chosen. It is fine-grained psamitic equigranular rock,
about 95% of which is made up of suboval quartz clasts. Other mineral grains are
present only as minorities (tourmaline, epidote, muscovite and zircone). Quartz
grains reach up to 0.1 mm in diameter, but those of muscovite are larger, up to
0.3 mm [3]. The matrix is formed by clay minerals (mainly kaolinite).
Also the argillite was very popular material in historical architecture. It was
used for sacral as well as for secural buildings, flagstone pavements, roof slabs,
and facing. The studied argillite is coming from quarry Dbn, Czech Republic.
Its main constituents are illite, calcite, minerals on the basis of SiO2 having
granularity 0.3 0.15 mm, feldspar, and mica, whereas rigid materials form 40
60% of argillite volume.
Within the reconstruction of historical buildings, renewal and restoration of
interior and exterior plasters50and
mmortars is often done. From the point of view of
a historian, it is not acceptable to use lime-cement mortars in Romanesque,
Gothic, Renaissance, and Baroque buildings. However, the pure lime-based
plaster does not exhibit sufficient resistivity against moisture action. On that
account, proper hydraulic admixtures must be used that enhance the durability of
mortars. These materials must have similar composition as the historical
materials and they have to be applicable by the original technological processes.
As the chemical analyses of many plasters from historical buildings show, the
past centuries external plasters that are preserved until today contain products
formed by lime reaction with pozzolanic or hydraulic admixtures. Pozzolanic
admixtures appeared to have positive effect on properties of lime binder in the
past. On that account, lime-based mortar with metakaolin addition as pozzolana
admixture was chosen. The composition of the lime-metakaolin plaster was as
3 Experimental methods
Basic material properties of all tested materials were determined at first. Bulk
density and matrix density were measured using gravimetric method and helium
pycnometry, and then total open porosity was calculated. The samples
dimensions for these measurements were 40 x 40 x 20 mm.
For determination of water vapour transmission properties we applied cup
method, according to the European standard EN ISO 12571 [5]. The
measurements were carried out in steady state under isothermal conditions. We
used dry cup arrangement of the experiment. Here, the sealed cup with the
studied material sample containing burnt CaCl2 (0% relative humidity) was
placed in a controlled climatic chamber at 250.5C and 50% relative humidity
and it was weighed periodically. The circular samples had diameter 95 mm and
thickness of 20 mm. The steady state values of mass gain were utilized for the
determination of the water vapour permeability. On the basis of measured water
vapour permeability, the water vapour diffusion coefficient and water vapour
resistance factor were calculated using the simple formulas given in [6, 7].
For determination of moisture diffusivity as function of moisture, the
moisture profiles were measured at first. The measurements were done on
samples having dimensions of 40 x 20 x 300 mm in 1-D experimental
arrangement of water transport. The moisture content in specific position in
sample was measured by capacitance technique calibrated by gravimetric
method. The moisture dependent moisture diffusivity was then calculated using
inverse analysis of measured moisture profiles by means of Boltzmann-Matano
treatment [8].
In the sorption isotherm measurement, the samples were placed into the
desiccators with different salt solutions to simulate different values of relative
humidity. The mass of samples was measured in specified periods of time until
steady state value of mass was achieved. Then, the volumetric moisture content
was calculated and sorption isotherm of each tested material was plotted.
The thermal conductivity as the main parameter of heat transport was
determined using the commercial device ISOMET 2104 (Applied Precision,
Ltd.). ISOMET 2104 is a multifunctional instrument for measuring thermophysical parameters which is based on the application of an impulse technique
and is equipped with various types of optional probes. Thermal conductivity was
measured in the moisture range from the dry state to full water saturation on the
70 mm cubes.
Materials Characterisation V
147
(in our case stone masonry wall) can be considered basically as a mixture of
walling blocks and mortar. In more precise calculations, each of these materials
can be further considered as a mixture of three phases, namely solid, liquid and
gaseous phase (in four phase systems, the effect of bound water can be included)
that form their matrix and porous space. There are two basic approaches that can
be applied for determination of thermal and hygric properties of the stone
masonry. The first possibility is to apply homogenization techniques on the
moisture dependent material data of the materials involved in the masonry. This
simplified procedure was used in this work. The second possibility is based on
the complete knowledge of material properties of the particular components
forming the porous body of the structure (dry stone, dry mortar, free water,
bound water and air). From the properties of particular components and their
volumetric fractions the effective properties of the stone masonry can be
accessed.
On the basis of previous experience with application of homogenization
techniques, we have chosen for the calculations performed in the presented work
original Lichteneckers formula that was proved to be satisfactory for evaluation
of moisture dependent thermal and hygric properties of porous building materials
[9]. Also the application of four phase models looks promising. However, in case
of the masonry studied in this work, determination of amount of bound water
represents very complex problem.
The Lichteneckers equation adjusted for the studied masonry wall assumes
that the effective hygric and thermal parameter of the considered material
satisfies equation
peff f b pb f m pm ,
k
(1)
peff
1
n
i 1
fi
pi
(2)
peff f i pi ,
(3)
i 1
where Eq. (2) represents the lower limit and (3) the upper limit of the
investigated effective material parameter (fi is the volumetric fraction of the
particular phase, in our case argillite, sandstone and mortar), pi its material
parameter).
5 Studied masonry
For application of homogenization theory for the evaluation of the effective
hygric and thermal parameters of masonry, the typical fragment of the wall was
constructed in a simplified way similar to common brick masonry as shown in
Figure 1.
Figure 1:
Using this scheme, the volumetric fractions of the particular materials in the
wall were calculated. The volumetric fraction of walling materials is equal to
0.824, whereas the volumetric participation of mortar is 0.176. Within the
calculations, effective parameters of two different walls were accessed. The
studied walls consisted always of lime based metakaolin mortar, and of one
walling material, nominally sandstone or argillite.
Materials Characterisation V
149
proved high porosity, what is very positive factor from the point of view of their
presumed application in reconstruction of historical buildings.
Water vapour transmission properties of masonry materials are presented in
Table 2. We can see systematically very low values of water vapour resistance
factor what is again very prospective finding for application of all tested
materials in historical masonry. Since the historical masonry usually exhibits
increased moisture content, there is necessary to apply within the reconstruction
processes such materials that will allow moisture evaporation from the renovated
structures within the warm periods of the year.
Table 1:
material
sandstone
argillite
lime-metakaolin
mortar
Table 2:
matrix density
(kg/m3)
2627
2235
2620
total open
porosity (-)
0.31
0.39
0.35
material
sandstone
argillite
lime-metakaolin
mortar
water vapour
water vapour
diffusion
resistance
coefficient (m2/s)
factor (-)
2.29E-11
3.11E-06
7.40
2.24E-11
3.05E-06
7.55
2.27E-11
3.08E-06
7.47
Argillite wall homogenization, wet cup arrangement 0-50% RH
water vapour
permeability (s)
water vapour
water vapour
diffusion
resistance
coefficient (m2/s)
factor (-)
2.74E-11
3.73E-06
6.17
2.61E-11
3.55E-06
6.47
2.69E-11
3.65E-06
6.30
Figure 2:
Materials Characterisation V
Figure 3:
Figure 4:
151
Figure 5:
Measured and calculated thermal conductivity of all studied materials and two
types of masonry is presented in Figures 6 and 7. The values of effective thermal
conductivity in dependence on moisture content are systematically between the
range of Wieners bounds, which basically proves reliability of applied
homogenization technique.
Figure 6:
Materials Characterisation V
Figure 7:
153
7 Conclusions
An example of application of homogenization technique for determination of
hygric and thermal parameters of the stone masonry was introduced in the paper.
The obtained results indicate that the Lichteneckers equation may be
successfully used for such type of applications. Nevertheless, there is still an
open task to verify the reliability of applied homogenization model by laboratory
experiments that should confirm or controvert the presented results.
Acknowledgement
This research has been supported by the Czech Ministry of Education, Youth and
Sports, under project No MSM 6840770031.
References
[1] Pavlkov, M., Pavlk, Z. & ern, R., Hygric and Thermal Properties of
Materials Used in Historical Masonry. Proc. of the 8th Symposium on
Building Physics in the Nordic Countries, Technical University of Denmark:
Lyngby, pp. 903-910, 2008.
[2] Vejmelkov, E., Pavlk, Z., Fiala, L., Pavlkov, M. & ern, R., Heat and
Moisture Transport Properties of Stone Masonry Materials. Proc. of the
Building Physics Symposium, Katolieke Universiteit Leuven, Leuven, pp.
113-116, 2008.
Materials Characterisation V
155
Abstract
SOFC (Solid Oxide Fuel Cells) appear to be a great alternative way to produce
electricity from hydrogen with high efficiency and no greenhouse gas emissions.
SOFC are efficient at high temperatures (around 800C) and are meant to be
used for stationary applications as heat and electric co-generation devices. To
understand how gases, electricity and heat flow through these media and to
improve their efficiency, it is critical to know the actual microstructure of these
electrodes. The studied electrodes are Ni-YSZ cermets in which characteristic
element sizes are around 1m. The 3D microstructure has been reconstructed
using FIB (Focused Ion Beam) tomography. This technique has been used on
several samples (different Ni-YSZ proportions) and gives representative 3D
volumes of around 101010m with a 10nm resolution. Theses 3D volumes are
then analysed to extract some important structural parameters such as volumetric
proportions, active surfaces, connectivity of each components and three phase
boundaries (TPB). Then, the 3D reconstructed volumes have been used to
determine homogeneous media equivalent properties such as thermal, ionic and
electric conductivities. These homogeneous equivalent properties are estimated
using a hot guarded plate simulation that takes into account each component
properties and the 3D structure.
Keywords: SOFC, FIB, tomography, thermal conductivity.
1 Introduction
Solid oxide fuel cell (SOFC) based-technology is one of the most promising
energy conversion systems due to its high efficiency and fuel flexibility [1]. The
materials involved in their constitution have become increasingly sophisticated,
both in composition and microstructure. One of the most common anode
WIT Transactions on Engineering Sciences, Vol 72, 2011 WIT Press
www.witpress.com, ISSN 1743-3533 (on-line)
doi:10.2495/MC110141
Experimental
Figure 1:
Materials Characterisation V
157
Figure 2:
Figure 3:
Materials Characterisation V
159
Figure 4:
Figure 5:
The TPB are the regions where the three phases (pores, 8YSZ, Ni) meet each
other. The TPB analysis procedure has been apply first to the initial volume and
yields to a TPB density of 11.2m.m-3. The activity of a TPB is subject to
numerous definitions in the literature. The major part of the authors defined a
TPB as active, if it lies three percolated phases (pore, Ni, 8YSZ) [4, 7]. The
active TPBL has been calculated and equal to 7.4m.m-3. This value is
relatively high compared to those reported in the literature. This suggests a high
electrochemical performance for the studied electrode.
WIT Transactions on Engineering Sciences, Vol 72, 2011 WIT Press
www.witpress.com, ISSN 1743-3533 (on-line)
Materials Characterisation V
161
. D 0
(1)
D is defined as the diffusion coefficient in the material. Each voxel is
considered uniform with a given diffusion coefficient depending of its phase
(Dvoxel. Table 1). This equation is solved using an implicit finite difference
method based on a conjugate gradient algorithm [14]. Different Dirichlet
boundary conditions are fixed on two opposite faces ( 1 on one face and
the other faces. Once the distribution within the volume is obtained
(Figure 6), the average flux () over a cross section of surface area S can be
calculated:
D voxel . n dS
(2)
S
Table 1:
Figure 6:
-1
Pores
(H2) 0.48
0
0
8YSZ
2
14
10-6
Ni
50
10-5
104
Deff S
1 2
(3)
xi max
The combination of eqns (2) and (3) leads to estimate Deff as:
Deff
Dvoxel
xi
S 1 2 S
xi max
dS
(4)
At 1123K and under one atmosphere the mean free path of H2 is around
900nm, and is bigger than our voxel size [15]. Then the concentration cannot be
considered uniform in a voxel and the bulk diffusion assumption is not valid
anymore. In the case of electron, ion and heat transport the corresponding mean
free path are smaller than 100nm. So, only heat, ion and electron diffusive
transport have been calculated in this work.
The results are given in Table 2 where the different values of Deff are reported
under the form of thermal, electronic and ionic effective conductivities. The
effective thermal conductivity appears quite isotropic with a slighter lower value
in the z direction, close to 4 W.m-1.K-1. This value has the same order than those
generally reported for SOFC anodes [16].
Table 2:
-1
x
4.23
1.43
281
y
4.54
1.19
252
z
3.27
1.03
66
4 Conclusion
This work has focused on the 3D analysis of the microstructure in a Ni-8YSZ
SOFC anode reconstructed by FIB tomography. Sample preparation, milling and
WIT Transactions on Engineering Sciences, Vol 72, 2011 WIT Press
www.witpress.com, ISSN 1743-3533 (on-line)
Materials Characterisation V
163
imaging conditions and data processing have been optimized to obtain a high
quality representative 3D microstructure having the following dimension of
8.669.79 11.41m3. Home made programs have led to the calculation of some
microstructural parameters such as volume fraction, connectivity, specific
surface and interface areas and TPB length including the proportion of active and
non active ones. These calculations have shown that for an anode with an initial
8YSZ/NiO volume ration of 60/40 leading after reduction to volume relative
amounts 41% of pores, 33% of 8YSZ and 26% of Ni, all the phases were highly
percolated. The active TPB length has been estimated as 7.4m/m3 which
comparing to the literature results corresponds to an effective anode. By solving
the diffusive transport equation with finite difference calculations, the effective
thermal, electronic and ionic conductivities could be determined.
References
[1] S. Singhal, K. Kendall, High-Temperature Solid Oxide Fuel Cells:
Fundamentals, Design and Applications, Elsevier Advanced Technology,
Oxford, UK, 2003.
[2] P.R. Shearing, J. Gelb, N.P. Brandon, X-ray nano computerised
tomography of SOFC electrodes using a focused ion beam samplepreparation technique, J. of the Europ. Ceram. Soc., 30 (2010) 1809-1814.
[3] K.N. Grew, A.A. Peracchio, W.K.S. Chiu, Characterization and analysis
methods for the examination of the heterogeneous solid oxide fuel cell
electrode microstructure, Part 2 : Quantitative measurement of the
microstructure and contributions to transport losses, J. Power Sources, 195
(24) (2010) 7943-7958.
[4] J.R. Wilson, W. Kobsiriphat, R. Mendoza, H.-Y. Chen, J.M. Hiller, D.J.
Miller, K. Thornton, P.W. Voorhees, S.B. Alder, S.A. Barnett, Threedimensional reconstruction of a solid-oxide fuel-cell anode, Nature
Materials, 5 (2006) 541-544.
[5] S.A.Barnett, J.R. Wilson, W. Kobsiriphat, H.-Y. Chen, R. Mendoza, J.M.
Hiller, D.J. Miller, K. Thornton, P.W. Voorhees, S.B. Alder, Three
dimensional analysis of solid oxide fuel cells using Focused ion beamScanning electron microscopy, Microsc. Microanal., 13 (2007) 596-597.
[6] D. Gostovic, J.R. Smith, D.P. Kundinger, K.S. Jones, E.D. Washman,
Three-dimensional reconstruction of porous LSCF cathodes, Electrochem.
Solid-State Lett., 10 (12) (2007) B214-B217.
[7] L. Holzer, B. Munch, B. Iwanschitz, M. Cantoni, T. Hocker, Th. Graule,
Quantitative relationships between composition, particle size, triple phase
boundary length and surface area in Ni-cermet for Solid Oxide Fuel Cells,
J. Power Sources, doi : 10.1016/j.jpowsour.2010.08.006.
[8] L. Holzer, B. Muench, M. Wegmann, P.H. Gasser, R.J. Flatt,
FIB-nanotomography of particulate systems Part 1 : Particle shape and
topology of interfaces, J. Am. Ceram. Soc., 89 (2006) 2577-2585.
[9] J.R Wilson, J.S Cronin, S.A. Barnett, Linking the microstructure,
performance, and durability of Ni-Yttria-Stabilized Zirconia solid oxide
WIT Transactions on Engineering Sciences, Vol 72, 2011 WIT Press
www.witpress.com, ISSN 1743-3533 (on-line)
[10]
[11]
[12]
[13]
[14]
[15]
[16]
[17]
[18]
[19]
[20]
[21]
fuel cell anodes using three dimensional focused ion beam Scanning
electron
microscopy
imaging,
Scripta
Materialia,
doi
10.1016/j.scriptamat.2010.09.025.
B. Ruger, J. Joos, A. Weber, T. Carraro, E. Ivers-Tiffe, 3D-modeling and
performance evaluation of mixed conducting (MIEC) cathodes, ECS
Transactions, 25 (2) (2009) 1211-1220.
<http://www.vsg3d.com>
J. Hoshen, R. Kopelman, Percolation and cluster distribution. I. Cluster
multiple labelling technique and critical concentration algorithm, Phys.
Rev. B., 1 (14) (1976) 3438-3445.
Y. Nakashima, S. Kamiya, Mathematica programs for the analysis of
three-dimensional pore connectivity and anisotropic tortuosity of porous
rocks using X-ray computed tomography image data, J. Nucl. Sci. Technol.,
44 (9) (2007) 1233-1247.
D. Rochais, G. Le Meur, V. Basini, G. Domingues, Microscopic thermal
characterization of HTR particle layers, Nucl. Eng. And Design, 238 (2008)
3047-3059.
T.G. Sherwood, R.L Pigford, C.R. Wilke, Mass transfer, In:Clark,
B.J. Maisel, J.W. (Eds.) McGraw-Hill Inc., New York, pp. 39-43, 1975.
S. Kaka, A. Pramuanjaroenkij, X.Y. Zhou, A review of numerical
modelling of solid oxide fuel cells, J. Hydrogen Energy, 32 (2007) 761-786.
D.R. Lide, Handbook of Chemistry and Physics, Thermal Conductivity of
gases, chap. 6, p 206, 89 th. ed., Taylor and Francis, Boca Raton (2008).
K.W. Schlichting, N.P. Padture, P.G. Klemens, Thermal conductivity of
dense and porous yttria-stabilized zirconia, J. Mat. Sci., 36 (2001)
3003-3010.
Y.S. Touloukian, R.W. Powell, C.Y.Ho, P.G. Klemens, Thermophysical
Properties of matter, vol. 1, Thermal conductivity, metallic elements and
alloys., IFI/Plenum, New-York-Washington, 1970.
R. Landauer, The electrical resistance of binary metallic mixtures, J. Appl.
Phys., 23 (1952) 779-784.
Z. Wu, M. Liu, Modelling of ambipolar transport properties of composite
mixed ionic-electronic conductors, Solid State Ionics, 93 (1997) 65-84.
Materials Characterisation V
165
Abstract
The aim of this work is to present a model of load transfer between porous
matrix and short fibres in ceramic matrix composites. This analysis is based on
the earlier shear-lag models used for polymeric composites. However, geometry
and strength of fibres in addition to the matrix porosity are included in the
present analysis. The theoretical curves for the longitudinal and shear stresses
distributions along the fibre-porous matrix interface are presented. They
exhibited a maximum strength point at the middle of the short fibres. It became
evident that the critical length is governed by the relative properties of the fibres,
matrix and porosity, which greatly influenced the load carrying capacity of the
fibres in the composites. In addition, the present simplified solution facilitates
the understanding of the interface mechanism using porous matrix.
Keywords: modelling, ceramic matrix composites, shear-lag models, porous
ceramics.
1 Introduction
Modern structural ceramic composites possess a number of unique properties
that cannot be achieved by other materials. Therefore, they have a potential for
saving energy, reducing wear, and increasing the lifetime of components [1].
However, regardless of their remarkable properties, structural ceramics are not as
widely used in industry as they should and could be. Among the reasons for the
reluctance of industry to introduce structural ceramic as components are [1]:
high price of ceramic parts,
insufficient knowledge in traditional metal-oriented engineering,
WIT Transactions on Engineering Sciences, Vol 72, 2011 WIT Press
www.witpress.com, ISSN 1743-3533 (on-line)
doi:10.2495/MC110151
2 Modelling
2.1 Previous considerations and analysis
Considering a loaded composite made of a dense fibre with length 2L embedded in
a porous matrix made of the same material of the fibre, hereby is assumed that no
slippage occurs between fibre and matrix. It should be also considered that the
Poissons ratio of fibre and matrix is the same, which implies on the inexistence of
transversal stress when the loading is applied along the fibre. Considering the
displacements in the fibre u and far away from the fibre v: (Fig. 1):
Figure 1:
Simplified scheme of the stress field around the fibre. (a) without
loading; (b) loaded [6].
Materials Characterisation V
167
E
A
L
(1)
dP EA d
dx
L dx
(2)
dPf
dx
B (u v )
(3)
where Pf is the load acting on the fibre and B is a constant that depends on the
fibre distribution and the Youngs Modulus of fibre and matrix.
Differentiation of Eq. (3) leads to:
d 2 Pf
dx
du dv
B
dx dx
(4)
The derivatives of u and v can be taken as the deformations in the fibre and
matrix, respectively:
Pf
du
dx Af E f
(5)
dv
dx
(6)
d 2 Pf
dx
B f
Af E f
(7)
Pf E f A f S sinh x T cosh x
where:
(8)
B
E f Af
(9)
Property
Fibre volume fraction
Matrix porosity fraction
Fibre length 2l (mm)
Fibre diameter (m)
Ratio Critical length/Length or
Value
0.45
0.24
50.8
10
0.25
Figure 2:
Materials Characterisation V
169
It can be proposed that the stress distribution between the points L-L and L
follows a behaviour such as:
f ( x ) Ax B
(10)
By using the boundary conditions given in Fig. 2, and substituting them in (10):
E f A( L L) B
(11)
(1 ) E f AL B
(12)
(1 ) E f AL E f A( L L)
(13)
E f A L
(14)
And then:
Ef
L
(15)
(1 ) E f
Ef
B
(16)
B 1 1 E f
(17)
f ( x)
Ef
1
x 1 1 E f
L
(18)
To determine the shear stresses along the fibre, the force equilibrium in a
fibre element with diameter 2r and length dx is made in the x direction, resulting
in:
(19)
r 2 2 rdx 0
Then, the shear stresses are given by:
r d
2 dx
(20)
f ( x)
Figure 3:
r E f
2 L
(21)
Figure 4:
With the stress distribution along the fibre, it is possible to calculate the
average stress carried by the fibre in the composite, given by:
1 L
f ( x )dx
L 0
(22)
1 L Ef
1
x 1 1 E f dx
0
L
L
(23)
Materials Characterisation V
Figure 5:
171
Then,
Ef
f 1
2
(24)
And for 0 < < 1, i.e., the fibre is longer than the critical length:
L L
E f dx
L L
f ( x )dx
(25)
Therefore:
f E f 1
(26)
With the average stresses well defined, we can define the stresses in the ply
longitudinal and transversal directions. When the matrix material is the same as
the fibre, it is possible to write the elastic modulus of the matrix as a function of
the fibre modulus:
Em E f e bp
where b is a shape factor that depends on the pore shape and distribution.
(27)
Figure 6:
T m E f e bp
(28)
The stress on the longitudinal direction is given by the average value between
matrix and fibre, based on the volumetric fractions of fibre and matrix:
L m 1 f f f 1 f E f e bp f f
(29)
L 1 f E f e bp E f 1
f
2
(30)
L 1 f E f e bp E f 1
f
2
(31)
Materials Characterisation V
Figure 7:
Figure 8:
173
3 Conclusions
A load transfer model was proposed to ceramic composites, which relates the
matrix porosity and fibre length with the mechanical strength of such
composites. It is hoped that the experimental validation of the model can be
established or any changes aimed at increasing its accuracy can be made.
Acknowledgements
The authors would like to thank the agencies CAPES (Coordenao de
Aperfeioamento de Pessoal de Nvel Superior), CNPq (Conselho Nacional de
Desenvolvimento Cientfico e Tecnolgico) and DFG (Deutsche
Forschungsgemeinschaft), for funding this research under grant Bragecrim
015/09.
References
[1] Janssen, R., Scheppokat, S. & Claussen, N., Tailor-made ceramic-based
components Advantages by reactive processing and advanced shaping
techniques. Journal of the European Ceramic Society, 28, pp. 1369-1379,
2008.
[2] Janssen, R. & Hotza, D. Low-cost and reliable production of oxide ceramic
matrix composites, Bragecrim Project Proposal, December 2008.
[3] Wendorff, J., Janssen, R. & Claussen, N., Saphirfaserverstrkung
reaktionsgebundener
oxider
Keramiken.
Verbundwerkstoffe
und
Werkstoffverbunde, ed. G. Ziegler, pp. 421-424, 1996.
[4] Lundberg, R. & Eckerbom, L., Design and processing of Al-oxide
composites. Ceramic Transactions, 58, pp. 95-104, 1995.
[5] Levi, C.G., Yang, J.Y., Dalgleish, B.J., Zok, F.W. & Evans, A.G.,
Processing and performance of an all-oxide ceramic composite. Journal of
the American Ceramic Society, 81, pp. 2077-2086, 1998.
[6] Casaril, A.; Gomes, E.R.; Soares, M.R.; Fredel, M.C. & Al-Qureshi, H.A.,
Anlise micromecnica dos compsitos com fibras curtas e partculas.
Matria, 12(2), pp. 408-419, 2007.
[7] Cox, H.L., The elasticity and strength of paper and other fibrous materials.
British Journal of Applied Physics, 3, pp. 72-79, 1952
Materials Characterisation V
175
Abstract
This paper is concerned with the axial behavior of the RC columns. Stress-strain
relationships of experimentally tested RC columns under concentric loading are
compared with the predictions of the Koksal model. Moment-curvature analyses
of RC sections are also performed employing the same model in a self-developed
moment-curvature program for confined concrete. Results are compared with the
output of EXTRACT which uses the Mander concrete model.
Keywords: confinement, reinforced concrete columns, concentric loading,
stressstrain relations.
1 Introduction
Confining pressure on RC columns increases the concrete strength. Transverse
reinforcements such as steel stirrups, FRP wraps, and steel jackets create a triaxial compressive stress state producing confinement action around the concrete
core. Extensive research on the improvement of concrete confinement has been
carried out since the pioneering study of Richart et al. [1]. The constitutive
model for confined concrete plays an important role on the accuracy of the
moment-curvature curves to perform a reliable pushover analysis of RC frames.
There are some frequently cited models (Hognestad [2], Kent and Park [3],
Sheikh and Uzumeri [4], Mander et al. [5], Saatcioglu and Razvi [6]) to predict
WIT Transactions on Engineering Sciences, Vol 72, 2011 WIT Press
www.witpress.com, ISSN 1743-3533 (on-line)
doi:10.2495/MC110161
2 Experimental study
The square specimens in the test program have 200x200 mm cross-section
dimensions and 1000 mm height. The columns were tested in the structural
laboratory of Yldz Technical University [8]. Figure 1 shows the details of test
setup and instrumentation for the two specimens.
Figure 1:
Test setup and details of test specimens C1L4S8 and C1L8S8 [8].
The experimental study was limited to one type of concrete mix design
typically. C1 type columns are tested at 30 days. All longitudinal bars are 10mm
in diameter and L4 and L8 shows the number of the bars in the cross-section of a
column. The tie spacing is 100mm and the S8 represents diameter of stirrups. For
measurement of axial strains, four linear variable displacement transducers
(LVDTs) are placed over the central 400mm gage length at each side of a
column in a similar way used to assess any eccentricity of the applied load as
recommended in the study of Shrive et al. [10]. Although the LVDT readings
were provided very close to each other, next to the maximum axial load there can
be a significant variation between the minimum and maximum values of
shortening reaching very high values [9]. Figure 2 shows the axial load-axial
WIT Transactions on Engineering Sciences, Vol 72, 2011 WIT Press
www.witpress.com, ISSN 1743-3533 (on-line)
Materials Characterisation V
177
shortening curve for C1L8S8 for which almost perfectly homogeneous strains
were obtained. However, for other columns the differences between LVDT
readings are somewhat greater indicating a localized damage zone at one specific
region of the specimen.
Figure 2:
(1)
fl
fl
fc
fc
(2)
1E0
E
1 0 2 1 1
cc cc
Es
(4)
cc ' c 1 k 2
fl
f 'c
(5)
where c' is the peak strain at the strength of plain concrete cylinders. In this
equation, k2 is taken as 5k1.
Figure 4 aims to compare experimental axial stress-strain curves of the four
column specimens with those predicted from Eqs (1)-(4).
In this study, a simple program is written in FORTRAN to produce momentcurvature plots for confined column section under bending and axial
compression. The cross section is divided into 1mm thickness slices. For a given
depth of natural axis, the strain at the extreme slice in compression is found by
iterative procedure. For each slices, the stress and strain are calculated using
force equilibrium and compatibility requirements respectively. Stress acting on
the core concrete (cover concrete is neglected) calculated using Koksals model.
Koksals model has also been utilized through the 3D finite element analysis of
RC and FRP-confined concrete columns [11, 13] successively. Figure 3(a)
WIT Transactions on Engineering Sciences, Vol 72, 2011 WIT Press
www.witpress.com, ISSN 1743-3533 (on-line)
Materials Characterisation V
179
Figure 3:
Figure 4:
Figure 5:
30
Moment (KNm)
C1L4S8
20
10
Model L200
EXTRACT L200
0
0
Figure 6:
20
40
60
Curvature x 10e6 (rad/m)
80
100
Materials Characterisation V
181
Moment (KNm)
30
20
10
Model L400
EXTRACT L400
0
0
20
40
Curvature x 10e6 (rad/m)
60
Figure 7:
Figure 8:
4 Conclusion
This paper has been primarily concerned with the modeling aspects for the axial
behavior of RC columns. Experimental axial stress-strain curves of the four
column specimens are predicted employing Koksal model. Furthermore, a
spreadsheet-based program is constructed to produce moment-curvature plots for
the performance-based design engineering approach. The following conclusions
can be drawn based on the results of the analyses:
1. As can be seen in Figure 5, Koksal model successively predicts the axial
behavior of the four RC column specimens tested under concentric loading.
2. It can be easily observed that Koksal model results into somewhat higher
values for the moment carrying capacities than the outputs of EXTRACT
program which employs Mander model in Figure 68, but the general trend
of the curves are the same. The developed program gives 27kNm and 38kNm
for moment carrying capacities at the axial load levels of 400kN and 600kN
while EXTRACT program results are between 22kNm and 27kNm.
Acknowledgement
The support of Boazii University Research Fund (ref: research project 5232)
for this paper is gratefully acknowledged.
References
[1] Richart, F.E., Bradtzaeg, A. and Brown, R. L. 1928. A study of the Failure
of Concrete under Combined Compressive Stresses. Bulletin Np. 185,
Engineering experimental station University of Illinois, Urbana, pp. 104.
[2] Hognestad, E., 1951. A Study of Combined Bending and Axial Load in
Reinforced Concrete Members. Bulletin Series No.399, University of
Illinois Eng. Exp. Station, Urbana.
[3] Kent, D.C. and Park, R. 1971. Flexural Members with Confined Concrete.
Journal of the Structural Division, Proc. of the American Society of Civil
Engineers, 97(ST7), pp.1969-1990.
[4] Sheikh, S.A. and Uzumeri, S.M. 1982. Analytical Model for Concrete
Confinement in Tied Columns. Journal of the Structural Division, Proc. of
the American Society of Civil Engineers, 108(ST12), pp. 2703-2722.
[5] Mander, J.B., Priestly, M.J.N. and Park, R. 1988. Theoretical Stress-Strain
Model for Confined Concrete. Journal of the Structural Engineering,
ASCE, 114(8), pp.1804-1826.
[6] Saatcioglu, M. and Razvi, S.R. 1992. Strength and Ductility of Confined
Concrete. Journal of the Structural Engineering, ASCE, 118(6),
pp.1590-1607.
WIT Transactions on Engineering Sciences, Vol 72, 2011 WIT Press
www.witpress.com, ISSN 1743-3533 (on-line)
Materials Characterisation V
183
[7] Kksal, H.O. 2006. A Failure Criterion for RC Members Under Triaxial
Compression. Structural Engineering and Mechanics, Techno-Press, 24(2),
pp.137-154.
[8] Turgay, T. CFRP Uygulanms_ betonarme elemanlarn performans (The
performance of FRP strengthened structural members). PhD thesis,
Submitted to Yldz Technical University; 2007.
[9] Kksal, H.O., Karako, C., Polat Z., Turgay T. and Akgn .
2007.Evaluation of Experimental Procedures for Confined Concrete
Columns. Computational Methods and Experimental Measurements XIII,
WIT Transactions on Modeling and Simulation, 46, 233-242.
[10] Shrive, P.L., Azarnejad, A., Tadros, G., McWhinnie, C. and Shrive, N.G.
2003. Strength of Concrete Columns with Carbon Fibre Reinforcement
Wrap. Canadian Journal of Civil Engineering, 30, pp. 543-554.
[11] Doran, B. 2009.Numerical simulation of conventional RC columns under
concentric loading. Material and Design, 30(6), 2158-2166.
[12] Saenz, L.P. 1964. Discussion of Equation for Stress-Strain Curve of
Concrete by Desai and Krishnan. ACI, 61(9), 1229-1235.
[13] Kksal, H.O, Doran, B., and Turgay, T. 2009. A practical approach for
modeling FRP wrapped concrete columns. Constr. and Build. Mat., 23(3),
14291437.
[14] EXTRACT-v.3.0.8, Cross section analysis program of structural engineers,
TRC/Imbsen Software Systems.
Section 5
Innovative experiments
Materials Characterisation V
187
Abstract
Composition and properties of the wood surface are very important in board
manufacture as they determine wood-adhesive bonding and the quality of the
final product. In this work two spectroscopic techniques, X-ray photoelectron
spectrometry (XPS) and time of flight secondary ion mass spectrometry (TOF
SIMS), have been employed to study the surface composition of eucalyptus
(Eucalyptus globulus) and ash (Fraxinus excelsior) sliced cut veneers. Both
wood species are widely used for decorative veneers in the finishing of wood
panels (particle boards, medium density fiberboards, etc.). Further
characterization of the wood surface was carried out by optic profilometry and
wettability analysis, using the sessile drop method to measure contact angles.
Wood is a very heterogeneous material and its composition can vary
significantly depending on the sampling area; then, the influence of the radial
position of the veneer in the trunk was also analyzed. From the low resolution
XPS spectra the oxygen to carbon (O/C) ratio was calculated and, resolving the
C1s signal, the percentages of the different carbon peaks corresponding to
different functional groups (C1, C2, C3 and C4 carbons) were also calculated.
The lower O/C ratio and the higher C1/C2 ratio for eucalyptus than for ash was
attributed to the higher concentration of extractives on the eucalyptus wood
surface that was confirmed by the TOF-SIMS spectra, which additionally
revealed a patchy distribution of the extractives. The higher hydrophobicity of
the eucalyptus wood surfaces was also supported by the results on the wetting
properties of the veneers. Eucalyptus veneers exhibited a significantly higher
constant wetting rate angle (cwra) and, consequently, a lower wettability than
WIT Transactions on Engineering Sciences, Vol 72, 2011 WIT Press
www.witpress.com, ISSN 1743-3533 (on-line)
doi:10.2495/MC110171
1 Introduction
Wood is a renewable resource composed mainly of three polymers, cellulose,
hemicelluloses and lignin, and various extractives such as fatty acids, sterols,
tannins, etc. Wood surface characteristics condition the interaction with
adhesives and binders in wood based products and, therefore, have influence in
the quality of the final product. Several methods have been proposed to
characterize material surfaces. Contact angle analysis is one the traditional
methods for the characterization of surfaces on a macroscopic scale, and
provides information on surface wetting properties [1].
Surface chemical composition of complex materials such as wood can be
determined using two complementary spectroscopic techniques, X-ray
photoelectron spectroscopy (XPS) and time-of-flight secondary ion mass
spectrometry (TOF-SIMS), the former generating more quantitative data and the
latter more qualitative data [2]. X-ray photoelectron spectroscopy (XPS) provides
the elemental composition for the outermost 510 nm surface layers and functional
groups, chemical bonding types, oxidation state, etc. can be additionally deduced
from XPS data. Time-of-flight secondary ion mass spectrometry (TOF-SIMS)
provides data for depths of 12 nm, including detailed chemical information via
identification of intact molecular ions or characteristic molecular fragments that are
emitted from the surface [3]. Additionally, it provides ion images, mapping the
lateral distribution of secondary ions signals from lignin, carbohydrates, extractives
and metals within an analyzed wood surface area [2, 4].
The aim of this paper was to evaluate and compare the surface properties and
composition of eucalyptus (Eucalyptus globulus) and ash (Fraxinus excelsior)
sliced cut veneers used as decorative veneers in the finishing of wood panels. A
wettability analysis, using the sessile drop method to measure contact angles,
was applied in combination with two spectroscopic techniques, X-ray
photoelectron spectrometry (XPS) and time of flight secondary ion mass
spectrometry (TOF SIMS). Roughness of the wood surface was analysed by
optic profilometry. Wood is a very heterogeneous material and its composition
can vary significantly depending on the sampling area; then, the influence of the
radial position of the veneer in the trunk was also analyzed.
2 Methods
2.1 Contact angle determination
Samples of eucalyptus (Eucalyptus globulus) and ash (Fraxinus excelsior) sliced
cut veneers with a humidity of 8% (wet basis) were used to prepare 2 cm (length)
WIT Transactions on Engineering Sciences, Vol 72, 2011 WIT Press
www.witpress.com, ISSN 1743-3533 (on-line)
Materials Characterisation V
189
Figure 1:
Materials Characterisation V
191
The values calculated for the initial (t=0; 0), final (t =240 s; f) and constant
wetting rate (cwra) angles are presented in Table 1. Not only initial contact
angles but also final and constant wetting rate angles were remarkably higher for
the eucalyptus veneers. Higher contact angles are indicative of a worse wetting
behaviour and could be related with a greater migration of hydrophobic
substances to the surface of eucalyptus veneers [7]. Differences between the 0,
f and cwra values of interior and exterior veneers were almost negligible which
confirms the previous qualitative observation.
Table 1:
Initial (0), final (f) and constant wetting rate (cwra) angles for
exterior and interior veneers of eucalyptus and ash wood with an
aqueous solution of chestnut shell extracts.
Veneer
Eucalyptus
Ash
0 ()
99.72
(5.83)
75.22
(1.44)
Exterior
f ()
cwra ()
45.25
50.8
(1.88)
(0.11)
28.01
27.99
(0.37)
(0.37)
0 ()
95.34
(4.77)
73.42
(9.91)
Interior
f ()
cwra ()
46.70
50.12
(6.77)
(0.10)
27.75
30.6
(8.43)
(0.76)
(Standard deviation)
Veneer
Exterior
eucalyptus
Interior
eucalyptus
Exterior ash
Interior ash
C
80.7
(2.2)
79.2
(0.1)
77.2
(1.0)
76.9
(1.4)
O
18.0
(2.3)
19.3
(0.7)
20.8
(1.1)
20.5
(1.2)
N
0.7
(0.2)
1.0
(0.4)
0.6
(0.1)
0.9
(0.1)
Si
0.6
(0.1)
0.8
(0.2)
1.2
(0.5)
1.2
(0.5)
Ca
0.3
(0.1)
0.5
(0.1)
(Standard deviation)
As seen in Table 2, carbon and oxygen are the main components in the wood
surfaces and small amounts of nitrogen and silicon were also detected in both
wood species, whereas calcium was present only in the ash samples.
Nitrogen is present in the form of amine/amide functional groups. This
assignment is as a result of the N1s peak position, which is at approximately 400
WIT Transactions on Engineering Sciences, Vol 72, 2011 WIT Press
www.witpress.com, ISSN 1743-3533 (on-line)
Figure 2:
XPS survey spectra (a) and high resolution C1s spectra (b) for
interior eucalyptus and ash veneers.
Materials Characterisation V
193
composition and the structure of wood components, carbon atoms bonded with
one oxygen atom (C2 component) should be greater. The observed increase in
the presence of C1 components, which was more significant for eucalyptus
samples, might be due to the migration of lipophilic extractives to the wood
surface [9].
Table 3:
Veneer
Exterior
eucalyptus
Interior
eucalyptus
Exterior ash
Interior ash
C1
66.5
(4.7)
65.4
(2.9)
60.1
(1.7)
60.0
(2.2)
C2
29.2
(3.4)
29.3
(3.2)
32.4
(0.4)
33.5
(2.4)
C3
2.3
(1.6)
2.8
(0.5)
5.0
(1.2)
4.1
(1.0)
C4
2.0
(0.4)
2.8
(0.3)
2.6
(0.1)
2.4
(0.7)
C1/C2
O/C
2.3
0.22
2.2
0.24
1.9
0.27
1.8
0.27
(Standard deviation)
This theory is reinforced by the values obtained for the O/C and C1/C2 ratios,
which are directly related to the chemical composition of wood constituents
(polysaccharides, lignin and extractives). The O/C ratio varied between 0.22 and
0.27 (Table 3). If experimental O/C ratio values are compared to those of pure
compounds, 0.83 for cellulose, 0.33 for lignin and 0.1 for extractives [10] the
preferential presence of lignin and extractives on the wood surface is
demonstrated, mainly in the case of eucalyptus veneers. Moreover, the C1/C2
ratio provides an additional evidence to support the above interpretations.
Theoretically calculated values for the C1/C2 ratio are equal to 0 for pure
cellulose, around 1 for lignin and 10 or higher for extractives [11]. Therefore,
values in the range from 1.8 to 2.3 confirm the presence of extractives on the
wood surfaces and especially on eucalyptus wood.
Finally, when comparing the mean values of the C1/C2 and O/C ratios and
the C1-carbon percentages, more hydrophobic material was found in exterior
eucalyptus veneers. The more hydrophobic character of the eucalyptus wood
samples is in agreement with the results obtained from contact angle
measurements which revealed better wetting properties for ash veneers.
3.3 Time of flight secondary ion mass spectrometry (TOFSIMS)
TOF-SIMS spectra were prepared to represent the masses from 1 to 700 u,
although the range from 100 to 200 u is the most interesting because in this
spectral region fragments of lignin and carbohydrates are detected. On the
contrary, in the region below 100 u, multiple fragments which are common to all
organic materials, the so-called non-specific organic fragments appear and,
therefore, fragments of this region of the spectrum are rarely used in the study of
WIT Transactions on Engineering Sciences, Vol 72, 2011 WIT Press
www.witpress.com, ISSN 1743-3533 (on-line)
Figure 3:
Materials Characterisation V
195
Figure 4:
Positive ion images for selected ions signals from eucalyptus and ash samples
are shown in Figure 5. Black colour indicates no ion signal and white colour
maximum signal intensity. Images of all the characteristic ions for lignin (137,
151,167 and 181 u) and extractives, mainly sterols (383, 397 and 425 u), are
presented together. Ion images showed that lignin and mainly extractives were
more prominent on eucalyptus than on ash veneer surfaces and, moreover,
revealed that both components were inhomogeneously distributed on the surface.
Figure 5:
Positive ion images for eucalyptus (a) and ash (b) samples.
Materials Characterisation V
Table 4:
197
Sample
Magnification
5X (1.2x0.9 mm2)
20X (298x227 m2 )
Eucalyptus
50X (119x91 m2)
STITCH (3x2 mm2)
5X (1.2x0.9 mm2)
20X (298x227 m2 )
Ash
50X (119x91 m2)
STITCH (3x2 mm2)
Ra (m)
11.31
(2.44)
3.44
(1.24)
2.30
(0.91)
15.91
(0.45)
12.17
(1.23)
8.78
(2.60)
6.10
(2.56)
27.38
(2.55)
Rq (m)
16.26
(4.63)
5.11
(0.66)
2.93
(1.00)
23.56
(4.70)
15.92
(1.23)
11.61
(4.14)
8.05
(3.54)
39.36
(6.80)
(Standard deviation)
Acknowledgements
This work was funded by Ministerio de Ciencia e Innovacin, FEDER Funds and
Plan E Fundy (CTQ2009-07539).
References
[1] Bryne, L.E. & Walinder, M.E.P., Ageing of modified wood. Part 1: Wetting
properties of acetylated, furfurylated, and thermally modified wood.
Holzforschung, 64, pp. 295-304, 2010.
[2] Englund F., Bryne L.E., Ernstsson M., Lausmaa J. & Walinder M.,
Spectroscopic studies of surface chemical composition and wettability of
modified wood. Wood Mater. Sci. Eng., 1-2, pp. 8085, 2009.
[3] Bryne, L.O., Lausmaa, J., Ernstsson, M., Englund, F. & Walinder, M.E.P.,
Ageing of modified wood. Part 2: Determination of surface composition of
acetylated, furfurylated, and thermally modified wood by XPS and ToFSIMS. Holzforschung, 64, pp. 305313, 2010.
[4] Tokareva, E.N., Fardim, P., Pranovich, A.V., Fagerholm, H.P., Daniel, G.
& Holmbom, B., Imaging of wood tissue by ToF-SIMS: Critical evaluation
and development of sample preparation techniques. App. Surf. Sci., 253, pp.
75697577, 2007.
[5] Vzquez, G., Gonzlez-Alvarez, J., Santos, J., Freire, M.S. & Antorrena,
G., Evaluation of potential applications for chestnut (Castanea sativa) shell
[6]
[7]
[8]
[9]
[10]
[11]
and eucalyptus (Eucalyptus globulus) bark extracts. Ind. Crops Prod., 29,
pp. 364-370, 2009.
Nussbaum, R.M., Natural surface inactivation of Scots pine and Norway
spruce evaluated by contact angle measurements. Holz Roh- Werkstoff, 57,
pp. 419-424, 1999.
Christiansen, A.W., Effect of ovendrying of yellow-poplar veneer on
physical properties and bonding. Holz Roh- Werkstof , 52, pp. 139-149,
1994
Popescu, C.N, Tibirna, C.N. & Vasile C., XPS characterization of naturally
aged wood, App. Surf. Sci., 256, pp. 13551360, 2009.
Sinn, G., Reiterer, A. & StanzlTschegg, S.E., Surface analysis of different
wood species using X-ray photoelectron spectroscopy (XPS). J. Mater. Sci.,
36, pp. 4673-4680, 2001.
Barry, A.O., Koran, Z. & Kaliaguine, S., Surface analysis by ESCA of
sulfite post-treated CTMP. J. Applied Polym. Sci., 39, pp. 3142, 1990.
Sernek, M., Kamke1, F.A. & Glasser, W.G., Comparative analysis of
inactivated wood surface, Holzforschung, 58, pp. 2231, 2004.
Materials Characterisation V
199
Abstract
A sophisticated semi-scale system which allows experimental simulation of heat
and moisture transport in building structures is employed for analyzing heat and
moisture transfer across interfaces in stone masonry formed by argillite walling
blocks and lime-pozzolana mortar. The dimensions of investigated structure are
close to a real wall but the measuring accuracy is the same as in a laboratory
experiment. Combined relative humidity and temperature mini-sensors are
utilized in continuous long-term monitoring. Measured temperature and relative
humidity profiles provide information on the properties of the interface between
the walling material and mortar which makes possible to identify the interface
permeability. This parameter can then be used in computational models of heat
and moisture transport in masonry which adds to the accuracy of model
predictions.
Keywords: heat and moisture transport, interface resistances, semi-scale
experiment, stone masonry, argillite, lime-pozzolana mortar.
1 Introduction
Computational modelling of coupled moisture and heat transport represents an
effective tool for prediction of behaviour of building structures exposed to
climatic load. For reliability of computational analysis, two requirements must
be met. At first, mathematical-physical model describing the coupled moisture
and heat transport must be formulated. From the theoretical point of view, this
model should be sufficiently sophisticated and complex, and should take into
account all the external and internal effects that affect the moisture and heat
WIT Transactions on Engineering Sciences, Vol 72, 2011 WIT Press
www.witpress.com, ISSN 1743-3533 (on-line)
doi:10.2495/MC110181
Materials Characterisation V
201
2 Semi-scale experiments
The semi-scale experiments are presently very popular in verification and
calibration of HAM (Heat, Air, and Moisture transport) models since they allow
monitoring of hygrothermal changes in the studied structures in more detailed
way and the certain specific cases can be studied what is not very common in the
case of full test house measurement [4, 5].
The semi-scale measuring system for determination of temperature and
moisture fields is designed in such a way that it simulates conditions, which are
as close as possible to the real conditions on building site, but it still maintains its
laboratory character, so that the expenses are kept considerably lower compared
to a real test house. Also the accuracy of applied measuring methods for
moisture content and temperature measurement is much higher compared to insitu measurements [6].
3 Experimental
Within the performed experiments, the moisture and heat transport in the fragment
of argillite wall was studied. Two separate semi-scale experiments were designed
and performed in order to evaluate the effect of interface resistances on the
transition between the argillite walling blocks and mortar. Within the first
experiment, the heat transport was simulated, whereas the relative humidity was
maintained on constant level. In the second experimental arrangement, relative
humidity transport was monitored at constant temperature conditions.
3.1 Studied masonry and inbuilt materials
Two materials typical for historical masonry in Central European territory were
used. The researched fragment of the stone masonry consisted of argillite blocks
WIT Transactions on Engineering Sciences, Vol 72, 2011 WIT Press
www.witpress.com, ISSN 1743-3533 (on-line)
sand 0 4 mm
(kg)
750
w/d ratio
(-)
0.23
bulk density
(kg/m3)
1688
1353
matrix density
(kg/m3)
2560
2235
Materials Characterisation V
Figure 1:
203
positioning into the tunnel between the climatic chambers, climatic loading
simulation, temperature and relative humidity monitoring, data evaluation. The
walling of the studied structure was done in the standard way using dry mortar
mix and wet technological process. The sensors for monitoring temperature and
relative humidity were placed to the investigated construction to beforehand
bored holes. The upper part of the bore opening was closed by silicon sealing.
Placing of the sensors was done regarding to the study of interface moisture and
heat transport (Figure 2).
For a proper setting of climatic conditions, it was necessary to achieve nearsteady-state conditions in the studied wall. Therefore, there was necessary to dry
freshly built wall to remove technological water. The drying was done at 50C
and 10% of relative humidity for one month.
After that, the first experiment simulating the heat transport at constant
moisture conditions was started. On one side of the studied wall, constant
temperature of -9.5C was maintained, whereas on the other side of the masonry
fragment, constant temperature of 25C was simulated. The relative humidity
varied during the experiment in the range of 10%, typically between 40 and
50% in respect to performance of climatic chambers. Duration of the first
experiment was 84 days.
Figure 2:
Materials Characterisation V
Figure 3:
Figure 4:
205
Figure 5:
Figure 6:
The relative profiles measured in the second semi-scale experiment are shown
in Figures 7 and 8.
Materials Characterisation V
207
Figure 7:
Figure 8:
transition
zone
Figure 9:
5 Conclusions
Experimental investigation of water vapour and heat transport across the
interfaces in stone masonry was done in the conditions of 1-D semi-scale
experiment. The climatic loading of the studied structure was chosen in such a
way that water vapour transport under isothermal conditions and heat transport at
constant relative humidity were simulated. The data obtained in both
experiments confirmed the significant effect of both interface and surface
resistances on water vapour and heat propagation, thus the necessity to include
these effects in computational modelling of coupled moisture and heat transport.
On the basis of measured data, determination of heat and water vapour
resistances between particular materials layers is possible using methods of
WIT Transactions on Engineering Sciences, Vol 72, 2011 WIT Press
www.witpress.com, ISSN 1743-3533 (on-line)
Materials Characterisation V
209
inverse analysis. Also, the surface resistances on the contact of studied building
structure with ambient air can be accessed. In this way, the results can contribute
to higher accuracy of future computational modelling. Neglecting the behaviour
of the transition zone between the particular materials forming stone masonry
structures, on the other hand, may lead to possible improper design of renovation
or construction solutions and subsequent material and structural damage.
Acknowledgement
This research has been supported by the Czech Ministry of Education, Youth and
Sports, under project No MSM 6840770031.
References
[1] Grunewald, J., Diffusiver und konvektiver Stoff- und Energietransport in
kapillarporsen Baustoffen, Ph.D. Thesis, TU Dresden, Dresden, 1997.
[2] ern, R. & Rovnankov, P., Transport Processes in Concrete, 1st ed.,
Spon Press: London, 547 pp., 2002.
[3] Pavlk, Z., Michlek, P., Pavlkov, M., Kopeck, I., Maxov, I. & ern,
R., Water and Salt Transport and Storage Properties of Men Sandstone.
Construction and Building Materials, 22(22), pp. 1736-1748, 2008.
[4] Krus, M., Rsler, D. & Sedlbauer, K., New model for the hygrothermal
calculation of condensate on the external building surface. Research in
Building Physics and Building Engineering, Fazio, Ge, Rao, Desmarais
(eds.), Taylor & Francis Group, London, pp. 329-333, 2006.
[5] Pavlk, Z., Mihulka, J., umr, J. & ern, R., Experimental monitoring of
moisture transfer across interfaces in brick masonry. Structural Faults and
Repair 2010 [CD-ROM], Engineering Technics Press Edinburgh,
Edinburgh, 2010.
[6] Pavlk, Z., Pavlk, J., Jiikov, M. & ern, R., System for Testing the
Hygrothermal Performance of Multi-Layered Building Envelopes. Journal
of Thermal Envelope & Building Science, 25(1), pp. 239-249, 2002.
Section 6
Mechanical
characterisation
and testing
Materials Characterisation V
213
Abstract
This paper investigates the tension/compression hardening behaviour of autobody steel sheets with the variation of the pre-strain and the strain rate. To
conduct tension/compression tests with the variation of the pre-strain and the
strain rate, an experimental method was established by using a newly developed
clamping device to suppress buckling of a specimen. The clamping device
provides the supporting force from compression-type coil springs during the test
with a conventional dynamic material fatigue testing machine. From
experiments, the tension/compression hardening behaviour was observed with
the variation of the pre-strain and the strain rate. Effects of the pre-strain and the
strain rate on the hardening behaviour were also investigated based on the
tension/compression test results.
Keywords: tension/compression test, clamping device, pre-strain, strain rate.
1 Introduction
Spring-back caused by the elastic recovery of the residual stress inside a formed
part has been one of the most significant sources of defects in the sheet metal
forming process in recent years. Spring-back predictability of numerical
simulation, however, has not been satisfactory because its performance is not
good enough to calculate accurate residual stress during the forming process. To
improve spring-back predictability in numerical simulation, many researchers
have been trying to use accurate information of the hardening behaviour of sheet
materials in finite element analysis. The hardening behaviour of sheet materials
2 Test preparation
To obtain a larger compressive strain range in the tension/compression test,
buckling of a specimen should be prevented by using a properly-designed
specimen and by imposing a sufficient clamping force. Fig. 1 shows three
representative buckling modes in the tension/compression test: (1) buckling in
the thickness direction in the gauge region (T-buckling); (2) buckling in the
Figure 1:
Materials Characterisation V
215
unclamped region (L-buckling); (3) buckling in the width direction in the gauge
region (W-buckling). T-buckling mode can be suppressed by imposing a
sufficient clamping force on a specimen with a special clamping device. Lbuckling and W-buckling modes can be prevented by designing an appropriate
specimen shape.
2.1 Specimen shape
Boger et al. [2] used the secant formula and the Euler method to calculate the
attainable compressive strain for L-buckling and W-buckling modes,
respectively. When the flow curve is expressed by the Swift model, the
attainable compressive strain for two buckling modes can be calculated as
follows:
1
B n
eL = e0 -1
W
(1)
eW =
p 2 nW 2
- e0
3G 2
(2)
where B and W are the specimen width in the gripping region and the gauge
region, respectively. G is the gauge length of the specimen. 0 and n are the
plastic strain for the yield stress and the hardening exponent in the Swift model,
respectively. If the gauge length of the specimen and the coefficients of the Swift
model are given, the attainable compressive strains are a function of the gauge
width. More detailed formulation procedure can be found in the reference [2].
The specimen design was performed using the above equations formulated for
the attainable compressive strain of L-buckling and W-buckling. The selected
steel sheets are SPCC and DP590 which are commonly used auto-body steel
sheets. Strains at uniform elongation before necking initiates are approximately
0.15 and 0.12 for SPCC and DP590, respectively. These values are assigned as
the maximum pre-strain for tension/compression tests without necking of a
specimen. Table 1 shows the coefficients of the Swift model and r-values for
SPCC and DP590 at a quasi-static state. A specimen shape for the uniaxial
tensile test [7, 8] was utilized as reference specimen dimensions for
tension/compression tests as shown in fig. 3(a). In order to reduce the required
clamping force to suppress T-buckling, the gauge length is first reduced from 30
mm to 20 mm. The small clamping force is favourable for the reliable
tension/compression test, which can reduce the frictional and biaxial effect on a
specimen. The attainable compressive strain was plotted with respect to the
gauge width as shown in fig. 2. From the plotted curves, the optimal gauge width
can be selected as 8.8 mm for SPCC and 8.2 mm for DP590, respectively. To
ensure a stable testing region for the two steel sheets, the final gauge width can
be determined to be 8.8 mm as shown in fig. 3(b).
r0
r45
r90
SPCC
555.772
0.01121
0.253
1.381
1.094
1.697
1.317
DP590
1079.261
0.00879
0.220
0.704
0.784
0.948
0.805
Material
0.4
L-buckling
SPCC
DP590
0.3
W-buckling
SPCC
DP590
0.2
SPCC, 0.15
DP590, 0.12
0.1
0.0
10
15
20
25
Figure 2:
(a)
(b)
Figure 3:
Materials Characterisation V
n +1
2u x
e
ln
1
G
3KGWt
F=
n +1
n
4(n + 1)d 2 ct
1
t
t
1
- G 2 + e0 m 2 d + 2 m 2 d + 2 tan (md )
217
(3)
2u x ( l - u x )
2p
,d=
G
2
u
p
1+ 2R
x
where t and ux are the thickness of a specimen and a certain edge displacement.
K, n and 0 are coefficients of the Swift model. For the tension/compression test,
the specimen thickness and width change in the gauge region during
deformation. These values can be calculated based on anisotropy of material.
where c =
1+ R
, m=
W = W0 -
t = t0 -
2u
W0
ln 1- x
1 + R
G
(4)
t0 R 2ux
ln 1
1 + R
G
(5)
The critical normal force defined in eqn. (3) is now used to calculate the
critical clamping force to suppress T-buckling in the tension/compression test.
More detailed formulation procedure of the critical normal force can be found in
the reference [6]. Fig. 4 shows the critical clamping force with respect to the
compressive plastic strain. From the plotted curves, the required clamping force
for SPCC and DP590 can be determined to be 2.003 kN and 2.275 kN,
respectively.
10
G20W8.8
SPCC
8
DP590
DP590
0.12
SPCC
0.15
6
4
2
0
0.00
0.05
0.10
0.15
0.20
Compressive strain
Figure 4:
Coil springs
Gap controlling bar
Bolt
Nut
Clamping plate
Compression plate
(a)
Figure 5:
(b)
3 Experiments
3.1 Test material
The tension/compression test of auto-body steel sheets was conducted with the
variation of the pre-strain and the strain rate. Test materials were SPCC and
DP590 with a thickness of 1.2 mm. The chemical composition of the sheet
metals is presented in table 2.
Table 2:
Material
SPCC
DP590
Mn
0.420
1.700
Si
0.040
0.123
P
0.0012
0.0162
S
0.0150
0.0009
Materials Characterisation V
219
Figure 6:
0.15
Actuator displacement: 3 mm
Strain rate: 0.001 /sec
0.10
0.05
SPCC
DP590
0.00
(a)
Figure 7:
(b)
Material
SPCC
DP590
0.001
O
O
O
O
O
O
O
1.0
X
X
O
O
X
X
O
Fd = Fm - Ff
(6)
Ff = m Fc
(7)
Materials Characterisation V
221
corrected curve by comparison with the acquired curve obtained from a tensile
test with unclamped.
In order to subtract the through-thickness stress, t , induced by the clamping
force, the von Mises yield function is utilized to calculate the effective stress
corrected from the biaxial effect.
1
2
(sm - st ) + sm2 + st2
s=
(8)
where m is the measured stress from the load cell. After both frictional and
biaxial corrections, the flow curve of the clamped specimen agrees well with that
of the unclamped specimen as shown in fig. 8.
400
SPCC
1000
300
200
Acquired curve (unclamped)
Acquired curve (clamped)
Corrected curve
100
0
0.00
0.05
0.10
500
800
DP590
Strain rate = 0.001 /sec
Clamping force = 2.275 kN
= 0.08
600
400
Acquired curve (unclamped)
Acquired curve (clamped)
Corrected curve
200
0
0.00
0.15
0.05
True strain
(a)
Figure 8:
250
Simple tension
Displacement (pre-strain)
1 mm (0.040)
2 mm (0.080)
3 mm (0.120)
4 mm (0.162)
0
-250
-500
0.00
0.05
0.10
0.15
250
Simple tension
Displacement (pre-strain)
1 mm (0.037)
2 mm (0.075)
3 mm (0.114)
4 mm (0.156)
0
-250
-500
0.00
0.20
0.05
True strain
Simple tension
Displacement (pre-strain)
1 mm (0.038)
2 mm (0.077)
3 mm (0.118)
4 mm (0.161)
0
-250
0.10
0.15
True strain
0.20
250
Simple tension
Displacement (pre-strain)
3 mm (0.116)
4 mm (0.159)
0
-250
-500
0.00
0.05
0.10
0.15
0.20
True strain
(c)
Figure 9:
0.20
0.05
0.15
(b)
500
250
-500
0.00
0.10
True strain
(a)
500
0.15
(b)
500
0.10
True strain
(d)
500
Simple tension
Displacement (pre-strain)
1 mm (0.032)
2 mm (0.070)
3 mm (0.112)
0
-500
-1000
0.00
0.05
0.10
1000
500
Simple tension
Displacement (pre-strain)
1 mm (0.034)
2 mm (0.073)
3 mm (0.118)
0
-500
-1000
0.00
0.15
0.05
True strain
(a)
DP590, 0.1 /sec
500
Simple tension
Displacement (pre-strain)
1 mm (0.029)
2 mm (0.066)
3 mm (0.109)
0
-500
-1000
0.00
0.05
0.10
500
Simple tension
Displacement (pre-strain)
3 mm (0.106)
0
-500
-1000
0.00
0.15
0.05
True strain
SPCC, 1 mm
250
0
-250
-500
0.00
0.05
0.10
0.15
SPCC, 2 mm
250
0
-250
-500
0.00
0.20
0.05
True strain
0
-250
0.10
0.15
0.20
250
0.05
0.20
SPCC, 4 mm
250
0
-250
-500
0.00
0.05
True strain
(c)
Figure 11:
0.15
(b)
500
SPCC, 3 mm
-500
0.00
0.10
True strain
(a)
500
0.15
(d)
500
0.10
True strain
(c)
Figure 10:
0.15
(b)
1000
True stress [MPa]
1000
0.10
True strain
0.10
0.15
0.20
True strain
(d)
Materials Characterisation V
1000
DP590, 1 mm
500
0
-500
-1000
0.00
0.05
0.10
1000
DP590, 2 mm
500
0
-500
-1000
0.00
0.15
0.05
True strain
0.10
0.15
True strain
(a)
(b)
1000
True stress [MPa]
223
DP590, 3 mm
500
0
-500
-1000
0.00
0.05
0.10
0.15
True strain
(c)
Figure 12:
4 Hardening behaviour
True stresstrue strain curves were obtained after frictional and biaxial
corrections, as shown in figs. 9 and 10. Hardening curves expand with the
increase of the pre-strain at various strain rates. To investigate the strain rate
effect on the hardening behaviour, true stresstrue strain curves were redrawn
with the variation of the strain rate as shown in figs. 11 and 12. Hardening curves
also expand with the increase of the strain rate at various pre-strains. It is well
known that the strain and the strain-rate hardening are caused by micro-structure
changes inside the materials. To confirm this general expectation, Huh et al. [9]
demonstrated by observing TEM experimental results that the mechanism of the
strain-rate hardening is the change of the dislocation structures and the increase of
the dislocation density. The mechanism of the hardening behaviour during
tension/compression loading also can be explained by the micro-structure change
of steel sheets caused by the pre-strain and the strain rate effect. From the
experiments, it is clear that the tension/compression hardening behaviour is
changed by the strain rate as well as by the pre-strain. This means that the prestrain and the strain rate effect on the tension/compression hardening behaviour
should be considered simultaneously to improve spring-back predictability in
numerical sheet metal forming simulation.
5 Conclusions
This paper investigates tension/compression hardening behaviour by performing
tension/compression tests using a newly developed spring-loaded clamping
device. Contributions in this paper are summarized as follows:
1) A specimen shape was designed with the secant formula and the Euler
method adopted by Boger et al. [2] to prevent L-buckling and W-buckling in the
desired tension/compression test range. The clamping force to suppress Tbuckling was calculated with the plate buckling theory. A spring-loaded
clamping device was newly developed to suppress T-buckling in the
tension/compression test. From experiments, it is noted that the dimension of a
specimen and the choice of compression-type coil springs are appropriate to
execute reliable tension/compression tests.
2) The strain in the gauge region of a clamped specimen was measured
directly by using a digital image processing technique in order to overcome the
inherent structural problems in the use of conventional contact-type
extensometers. The measuring scheme using the digital image processing
technique can provide accurate strain history in the gauge region of a clamped
specimen during tension/compression tests.
3) With the test conditions established, hardening curves were reliably
obtained for various pre-strains and strain rates. The tension/compression
hardening curves were expanded with the increase of the strain rate as well as
with the increase of the pre-strain. The mechanism of the pre-strain and strain
rate hardening was explained by micro-structure changes, such as dislocation
structures and dislocation density. Experiments provide good information of the
hardening behaviour with respect to the pre-strain and the strain rate.
4) The newly developed clamping device can be directly applied to
tension/compression tests at high strain rates because of its compactness and
convenience to control the clamping force with various tensile testing machines.
This device guarantees the response of the clamping pressure in real time when
the clamping condition is changed by the specimen deformation during
tension/compression tests.
References
[1] Lee, M. G., Kim, D., Kim, C., Wenner, M. L., Wagoner, R. H. & Chung, K.,
Springback evaluation of automotive sheets based on isotropickinematic
hardening laws and nonquadratic anisotropic yield functions, part II:
characterization of material properties. Int. J. Plast., 21, pp. 883914, 2005.
[2] Boger, R. K., Wagoner, R. H., Barlat, F., Lee, M.G. & Chung, K.,
Continuous, large strain, tensioncompression testing of sheet material. Int.
J. Plast., 21, pp. 23192343, 2005.
[3] Cao, J., Lee, W., Cheng, H. S., Seniw, M., Wang, H. P. & Chung, K.,
Experimental and numerical investigation of combined isotropickinematic
hardening behavior of sheet metals. Int. J. Plast., 25, pp. 942972, 2009.
Materials Characterisation V
225
[4] Lee, M. G., Kim, D., Kim, C., Wenner, M. L. & Chung, K., Springback
evaluation of automotive sheets based on isotropickinematic hardening
laws and nonquadratic anisotropic yield functions. part III: applications.
Int. J. Plast., 21, pp. 915953, 2005.
[5] Taherizadeh, A., Ghaei, A., Green, D. E. & Altenhof, W. J., Finite element
simulation of springback for a channel draw process with drawbead using
different hardening models. Int. J. Mech. Sci., 51, pp. 314325, 2009.
[6] Cao, J. & Wang, X., An analytical model for plate wrinkling under triaxial
loading and its application. Int. J. Mech. Sci., 42, pp. 617633, 2000.
[7] Huh, H., Kim, S. B., Song, J. H. & Lim, J. H., Dynamic tensile
characteristics of TRIPtype and DPtype steel sheets for an autobody. Int.
J. Mech. Sci., 50, pp. 918931, 2008.
[8] Huh, H., Lim, J. H. & Park, S. H., High speed tensile test of steel sheets for
the stressstrain curve at the intermediate strain rate, Int. J. Automotive
Technol., 10(2), pp. 195204, 2009.
[9] Huh, H., Yoon, J. H., Park, C. G., Kang, J. S., Huh, M. Y. & Kang, H. G,
Correlation of microscopic structures to the strain rate hardening of SPCC
steel. Int. J. Mech. Sci., 52, pp. 745753, 2010.
Materials Characterisation V
227
Abstract
Analytic formulas for averaged elastic and strength characteristics of folded
structures determination are identified. They are based on the introduction of a
hypothesis about momentless action of its elements (web covers) in a charge
operation and on phased deletion from action ex-post buckling. Dimensionless
coefficients contained in the structural formulas, which appear on the solution
process of formulated problems, are to be determined from experimental
evidence on tension, compression and pure shear in two-planes. In a tensioncompression process the folded structures should be considered heterogeneous
material with averaged plasto-elastic behaviors, which has relations between
averaged voltages and averaged appropriate deformations and has a deformations
stepwise change area (conditional plastic flow).
Keywords: folded structures, cell of cyclicity, averaged elastic and strength
characteristics, modulus of elasticity, deformation curve, structural formulas,
theoretical and experimental method.
Introduction
x01 z . Therefore,
r xk e1 yk e2 , r
k
k
k
xk e1 yk e2
(1.1)
ei ei 1, ei ei
k
1, e1 e2 e1 e2 cos k
k
(1.2)
Materials Characterisation V
229
(1.3)
ke2 k ke2 k 0
()
(b)
Figure 1.
Parameterizations (1.1) correspond to the basis vectors (for the elements
xk e1 , r2 r
k
yk e2
(1.4)
aiik
(1.5)
k
12
aii
cos
1
a
, a12k a21k
2 k
2
a k sin k
a k
sin k
k k
k k
(1.6)
k sin 2 k .
t 2 z k t 2
(1.7)
where
vectors
normal
to
the
plane
k ,
satisfying
the
equations
ei m 0, m m 1 .
k
and
k w,xk , k w,yk
where
, x
ij ij k z k ijk
zk
(1.8)
where
1
2
k
k
k
k
u, y v,x w,x w,y
1
2
212
k
k
11
w,xxk , 22k w,yyk , 12
w,xyk
(1.9)
(1.10)
ij
zk
ijk
Materials Characterisation V
ijk
E
z
Eijml
k
k ml
2
1 v
231
(1.11)
where E , v are the Youngs modulus of elasticity and Poisson's ratio of wall
ijml
Eijml
vaijk amlk
k
1 v im jl
a k a k ailk ajm
k
2
(1.12)
, E1122
E2211
,
k
k
k
k
2
sin k
sin 4 k
E1112
E1121
E1211
E2111
E1222
E2122
E2212
k
k
k
k
k
k
k
E2221
cos k
,
sin 4 k
E1212
E2121
E1221
E2112
k
k
k
k
1 cos 2 k v sin 2 k
2sin 4 k
(1.13)
We assume that under the loading of the core in the sandwich and multilayer
structural elements in its walls because of their small thickness realized the flat
and membranaceous stress-strain state. In this case, accumulated in them
potential energy of deformation is permissible to calculate by the following
formulas
l d
1k
k
(1.14)
sin k dxk dyk
k t 11k 11 k 2 12k 12 k 22k 22
200
where d k
xz , yz , z
1
xz xz yz yz z z dV
2
V
xz , yz
(1.15)
of
xz
u w
v w
w
, yz
, z
z x
z y
z
(1.16)
Using of the principle of equivalence of real and replacing conditional core with
average elastic and strength characteristics requires to preparation of the
equality.
(1.17)
11k , 22k
and
12k .
11k
and
22k ,
and
12k 0 ,
consideration pinning the edges of the walls to construct the solution of the
stability function w
can be represented as
k
w k Amn
sin
m xk
m yk
sin
lk
d
(2.1)
Materials Characterisation V
233
where m 1, 2,..., n 1, 2,... are the wave numbers. Then with the
introduction of the dimensionless parameters
lk
,
d
Kx
11
k
k 22k 11k .
Dk
2E t
k
K x Dk , Dk
sin 4 k
12 1 v 2 lk
(2.2)
where
k
Kx
k2 n 2 4m 2 n 2 k2 cos 2 k
2
m k n
2
2
k
, Dk
represented as
w k W k sin
In a case
(2.3)
11k 0 , 12k 0
22k 0
Dk
sin 4 k
11k 22k 0 ,
critical tension
12k
k 5.34 4
yk
lk
sin
and
can be
yk kk xk
(2.4)
lk2
2
Dk 1 cos k
d2
2
1 cos k
1 cos 2 k
(2.5)
z
(3.1)
u0 i v0 j w0 k , 0 z h
h
where u0 const , v0 const and w0 const moving the points of the
z h,
with
a
net
in-plane
shear
boundary
plane
x01 z u0 0, v0 0, w0 0 , y 01 z u0 0, v0 0, w0 0
and
tension-
k 1, 2,1, 2
(3.2)
U 0 k u0 k e1k v0 k e2k w0 k m k
approximations
x1 1 1 x1 1 2 x2 2
U 0 , U U 0 , U 1 U 0
l1
l1
l2
x 2
1 2 U 0
l2
1
U
(3.3)
The condition of equivalence of the real and the conventional core requires the
satisfaction of the equality
k
k
k
k
k
(3.4)
u0 i v0 j w0 k U 0 u0 e1k v0 e 2k w0 m k 1, 2,1, 2
(3.5)
u 1
x1 1 1 x1 1 1 x1 1 1 x1 1
u0 , v v0 , u u0 , v v0
l1
l1
l1
l1
x 2 2 x 2
2
u 1 2 u0 , v 1 2 v0
l2
l2
(3.6)
x
x
u 2 1 2 u0 2 , v 2 1 2 v0 2
l2
l2
By the substitution into the linear kinematic relations
(3.7)
compute the components of deformation in the walls of the core in the membrane
approximation.
WIT Transactions on Engineering Sciences, Vol 72, 2011 WIT Press
www.witpress.com, ISSN 1743-3533 (on-line)
Materials Characterisation V
235
Compiled on the basis of the relations, we consider the case of the periodicity
cell of core by tension and compression along the axis z w0 0, uo v0 0
in the case of loading we can get formulas:
11 k 11 k sin 2 k
w0 k
k
k
k
, 22 22
12 12 0
h
E
E sin 2 k w0
1111 k
E 11
1 v2 k
1 v 2 sin 4 k h
E
E2211
22k
11 k v sin 2 k cos 2 k 11k
k
2
1 v
2 E 1211 k
E 11 cos k 11k
12k
1 v2 k
(3.8)
11
k
(3.9)
w0
u
v
w
, xz 0 , yz 0 , 33 Ez 0
h
h
h
h
(3.10)
u0 23
v0
13
Gxz , G yz
h
h
where the homogenized elastic modules Ez , Gxz and G yz must be determined.
By using the relations (3.10) in accordance with expression (1.15) for calculation
we arrive at the formulas in the case u0 v0 0
(3.11)
(3.12)
2
0
and by using relations (3.8) and (3.9), in accordance with (1.14) for calculation
k we arrive to the formulas
sin k sin k
w0 k 1, 2,1, 2
2
(3.13)
sin 3 k w02
E
td
k 1, 2,1, 2
1 v 2 2sin 3 k h
(3.14)
11k td
Ez
2
sin 3 k
E
t
(3.15)
and using (3.13) from (1.17) we can obtain the following dependence
33
2
t
11k sin k sin k
(3.16)
four walls of the core begins. Under these assumptions to determine the limit
33
33
2
t
sin k sink
(3.17)
33
From it by the experimentally determined value determined the quantity
.
When compressing of the core begins its destruction due to the destruction of
the buckling of the wall, which formed the voltage
k v sin 2 k cos 2 k
11k
and cos k
at a given dimensions
12k 11k
11
k
Kx
t
2 E sin 2 k t
k
Kx
(3.18)
12 1 v 2 sin 4 k lk
12 1 v 2 sin 4 k h
2E
Assuming further that for the real core loss of stability of a wall immediately
causes a loss of stability of a second wall, we form the equation
11k 11k ,
2 t
w0
K
x
12 h
h
(3.19)
Materials Characterisation V
where K x K x
237
K x . However
2E
t
K x
2
12 1 v sin ctg1 ctg 2 h
33
k of
33 of
sin 3 k
sin
k 1
(3.20)
Conclusions
References
[1]
[2]
[3]
[4]
[5]
[6]
[7]
[8]
[9]
Noor A. K., Burton W.S., Bert Ch. W., Computational models for
sandwich panels and shells. Applied Mechanics Reviews, 1996, V. 49,
3, p. 155-199.
Bersudsky V.E., Krisin V.N., Forest S., Manufacturing technology of
cellular aircraft structures. - Moscow: Mashinostroenie, 1975. 216s.
Aleksandrov A.Y., Bryukker L.E., Kurshin L.M., The calculation of
sandwich panels / M. Oborongiz, 1960. 272s.
Panin V.F., Gladkov Yu.A., Constructions with cores. Directory.
Moscow: Mashinostroenie. 1991. 272s.
Kryutchenko V.E., Analysis of optimum insulation properties of
sandwich plates with honeycomb core. /Mechanics of Composite
Materials. - 1993, T.29.- 6.- S.835-839.
Relsey S., Gellatly H. and Clark B., The shear modulus of foil
honeycomb cores. Aircraft Engineering. 1958. V.30. 356. P.294-302.
Bakhvalov N.S., Ponasenko G.P., Averaging of the processes in periodic
media.- Moscow: Nauka, 1984. 352 c.
Sachenkov A.V., Theoretical and experimental method for studying the
stability of plates and shells. Investigations on the theory of plates and
shells. Kazan: Izdatel'stvo Kazan. State. Univ. - 1970, Vol. 6.7.- S. 391433.
Volmir AS Stability of elastic systems. M.: Gos. out of Sci. literature.
1963. 880 sec.
[11]
[12]
Akishev N.I., Zakirov I.I., Paimushin V.N., Shishov M.A., Theoreticalexperimental method for determining the averaged elastic and strength
characteristics of honeycomb of sandwich structures. Mechanics of
Composite Materials. - 2011 (in press).
Zakirov I.M., Alekseev, K.A., Akishev N.I. Kayumov R.A., NikitinA.V.,
Zakirov I.I., Manufacturing of sandwich panels with folded core of
polymer paper Kazan: Publishing house Fan 2009
Zakirov I.M., Alekseev K.A., Determining of the parameters of the four
radial helical folded structure / IVUZ Aviatcionnaya technika 2005
Materials Characterisation V
239
Abstract
Almost pure aluminum composite, P/M Al-special, with an average grain
diameter of 1m with oxides on the grain surface was produced by a powder
metallurgy route. Billets were extruded on an indirect press at various
temperatures and ram-speeds. Specimens made from an extruded profile were
additionally annealed at 300C at various times. Hot workability of these various
initial states was studied by hot compression tests in the strain rate range 0.01
10 s-1, the temperature range 300580C up to a strain of 0.9. Additionally tensile
tests were carried out to determine the mechanical properties on specimens
previously annealed at 300C. The oxide surface prevents grains coarsening
during hot deformation and consequently excellent mechanical properties of the
specimens annealed at a temperature of 300C were obtained.
Keywords: P/M 1080 Al composite, hot extrusion, annealing, hot compression,
mechanical properties.
1 Introduction
Demands for weight reduction of parts applied in cars and aircrafts has lead to
the replacement of conventional materials with lighter and stronger materials.
Aluminum alloys have a great potential for this application but their weaknesses
are too low strength and stiffness. On the other hand production of parts from
metal-matrix composites which is based on the reduction of grain size is very
promising due to their unique mechanical properties in comparison with coarsegrained materials [16].
Powder (P/M) and ingot metallurgy which include particle reinforcement
mixing with solid or molten metal assure production of composites with
WIT Transactions on Engineering Sciences, Vol 72, 2011 WIT Press
www.witpress.com, ISSN 1743-3533 (on-line)
doi:10.2495/MC110211
2 Experimental procedures
2.1 Material and characterization of microstructure
A JSM-5610 energy dispersive spectrometer from iXRF Systems Inc. with a
digital processor 500 for chemical analysis of P/M Al-special was used along
Materials Characterisation V
241
Figure 1:
P/M Al-special
1st billet
2nd billet
3rd billet
500
500
500
360
360
360
418
340
320
8.87
6.2
3.89
7.06
4.94
3.10
50
50
50
2.4 Annealing
In order to investigate the influence of elevated temperature on recrystallization
resistance, annealing tests were carried out at temperatures of 300C and
employed at different times, i.e. 1, 5, 9 and 20 hours.
2.5 Hot compression and tensile tests
Hot workability was studied by hot compression tests. Cylindrical specimens of
Rastegew type with dimensions =10 mm x 15 mm were cut from the extruded
profile (3rd billet) so in non-annealed states as well as in annealed states, i.e.
previously annealed at 300C for 1, 5, 9 and 20 hours, respectively. Hot
compression of extruded material was carried out on a computer controlled
servo-hydraulic machine, Gleeble 1500D, in the temperature range 300C to
570C, at five different strain rates (0.01, 0.1, 1 and 5 or 10 s-1) up to a strain of
0.9, while for the annealed material from a production point of view only at a
WIT Transactions on Engineering Sciences, Vol 72, 2011 WIT Press
www.witpress.com, ISSN 1743-3533 (on-line)
Materials Characterisation V
243
strain rate range 0.15 s-1. The heating rate was 1C/s and the soaking time at
compression temperature was 1 minute. For reduction of friction between the
cylindrical specimen and the tool, and in order to avoid their mutual welding,
graphite lubricant and tantalum follies were used. After deformation the
specimens were water quenched. The conditions for hot compression testing are
given in Table 2.
For determination of tensile strength and elongation at room temperature
tensile tests were carried out on a Zwick Z400 according to standards (SIST EN
10002-1:2002, 5th edition, June 2002). Tensile specimens was cut out from the
centre of the extruded bar, the working length of the tensile specimens was 50
mm with a diameter of =10 mm. Tensile tests were also carried out also on a
AA 1050.
Table 2:
State of material
Extruded
Annealed
Strain range
0-0.9
0-0.9
Figure 2:
Figure 3:
Materials Characterisation V
Figure 4:
245
At higher strain rates the flow curves after achieving the maximal value at the
very beginning of hot compression, i.e. in the strain range up to 0.03, the values
of flow stress begin slightly to fall approaching a steady state (see fig. 5, right).
Moreover, at higher strain rates (110 s-1) the cracks did not occur that is
expressed also by the shape of flow curves.
Figure 5:
As mentioned the hot compression tests were carried out also for samples
annealed at various times. The shapes of flow curves obtained after various
annealing times and for a strain rate of 5 s-1 are similar to those obtained for the
extruded state but in general their values decrease up to an annealing time of 5
hours (see fig. 6). It is visible that values of flow curves are considerably lower
(about 15%) for annealed states in comparison to non-annealed states. During the
extrusion a small percentage of the oxide layer on the grains are torn. This may
allow re-arrangement of some grains with orientation changes. During annealing
of extruded material the process of part polygonization starts. After one hour at
300C material became stable since grain growth was blocked by oxides.
Figure 6:
Figure 7:
For comparison for AA 1050 the values for strength were around 80 MPa and
for elongation were around 38%. Thus around 2.6 times higher values for tensile
WIT Transactions on Engineering Sciences, Vol 72, 2011 WIT Press
www.witpress.com, ISSN 1743-3533 (on-line)
Materials Characterisation V
247
strength and 1.6 times lower values for elongation for P/M Al-special in
comparison to AA 1050 were obtained.
From the obtained results for P/M Al-special it can be derived that this
material can be applied also at elevated temperatures. It is also worth mentioning
that in our case grains consist only of almost pure Al and without particles in the
matrix. Thus the achieved mechanical properties are also on a relative high level.
4 Conclusions
New P/M Al-special composite made from almost pure Al with an average
diameter of 1m and with oxides on the grain surface was developed. The
achieved density after cold isostatic pressing was 84%, after hot extrusion 96%
and 100% after hot compression. The P/M Al-special material exhibited a good
combination of strength and elongation properties also at samples previously
exposed to elevated temperature.
The produced P/M Al-special has a great potential to substitute the classical
aluminum alloys and steels in the field of an elevated temperature working
environment. The obtained values for tensile strength for P/M Al-special alloy
are up to 2.6 times higher and for elongation around 1.6 times lower in
comparison to values obtained for AA 1050 aluminum alloy produced by
classical technology. The tensile strength of samples taken from extruded billet
with an extrusion ratio of 13.3 reaches a value of about 209 MPa while achieved
elongation was around 24%. The material exhibits good resistance against
increasing of grain size at elevated temperature. After 20 hours of annealing the
tensile strength slightly decreased and reached the value of 204 MPa. On the
other hand the value for elongation remains almost at the same level, i.e. 23%.
Extruded material exhibited low hot deformability at strain rates of 0.1 s-1 and
lower, while at higher strain rates the compressed samples were crack free.
Acknowledgements
The authors gratefully acknowledge Martin Balogand and Karol Idinsk from
Slovak Academy Of Sciences, Institute of Materials and Machine, Mechanics
Raianska 75, 831 02 Bratislava 3, for their contribution towards analyzing the
powder.
References
[1] Sahin, Y., Aclar, M., Production and properties of SiCp-reinforced
aluminium alloy composites, Composites Part A: Applied Science and
Manufacturing, 34(8),pp. 709-718, 2003.
[2] Shi, X.L., Mishra, R.S, Watson, T.J., Effect of temperature and strain rate
on tensile behavior of ultrafine-grained aluminum alloys, Materials Science
and Engineering: A, 494(1-2), pp. 247-252, 2008.
Materials Characterisation V
[17]
[18]
[19]
[20]
[21]
[22]
[23]
[24]
[25]
[26]
[27]
[28]
[29]
249
Materials Characterisation V
251
Abstract
Coarsening of precipitates in nickel-base superalloys is one of the mechanisms
by which creep damage occurs in these alloys. This is brought about by a
concomitant increase in the interparticle spacing, which results in faster
dislocation movement and therefore, faster creep strain accumulation. Therefore,
it is vital to accurately quantify the coarsening kinetics, which is essential for
creep prediction via microstructure-based continuum damage mechanics models.
In the present article, we report on coarsening studies of bimodally distributed
precipitates in DS GTD111TM (Trademark of the General Electric Company).
The baseline microstructure of GTD111TM consisted of a cuboidal secondary
population and a much finer, spherical tertiary population with respective
mean radii of 266 and 34nm. Long-term aging experiments were conducted on
baseline samples at temperatures of T1<T2<T3<T4 in the range of 800 to
1000C. At T1, there was no clear trend in coarsening of the secondary , likely
due to interference effects from the tertiary distribution. Therefore, secondary
coarsening was studied using data at higher temperatures where the distribution
was clearly unimodal. The kinetics was extracted assuming the cubic rate law to
be valid under conditions of volume diffusion controlled coarsening.
Subsequently, the secondary kinetics was applied to predict coarsening of the
tertiary distribution at T1 by incorporating a correction factor for volume fraction
WIT Transactions on Engineering Sciences, Vol 72, 2011 WIT Press
www.witpress.com, ISSN 1743-3533 (on-line)
doi:10.2495/MC110221
1 Introduction
GTD111TM is a directionally solidified nickel-base superalloy which is used in
hot gas-path sections of gas turbines at temperatures of 750 to 800C [1]. The
presence of a bimodal distribution of the (Ni3(Al,Ti)) precipitate phase
contributes to superior high temperature strength [2] of the alloy. However, upon
prolonged thermal exposure, the distribution undergoes coarsening [3],
resulting in an increase in the average interparticle spacing. This leads to easier
movement of dislocations in the alloy and consequently, lower creep resistance
[4]. In this context, coarsening as a damage mechanism warrants a detailed
study so that an understanding of the kinetics of microstructure degradation may
subsequently be used to develop continuum damage mechanics models.
Coarsening studies in multimodal precipitate systems has been reported in
literature in both blade alloys such as GTD111 [5] and IN738 [6] and disc alloys
such as Nimonic 115 [7] and Waspaloy [8]. Sharghi Moshatghin and Asgari [9]
studied coarsening of bimodal distribution in IN738LC in terms of a
compound size parameter that incorporates both primary and secondary
dimensions. They found that the kinetics analyzed using this size parameter
followed the cubic rate law in the range of 850 to 900C. Stevens and Flewitt [6]
reported volume diffusion controlled coarsening kinetics for both secondary and
tertiary in IN738 in the temperature range of 750 to 850C. Coakley et al. [7]
implemented the numerical mean-field model based on the LSW theory,
originally developed by Chen and Voorhees [10] to study bimodal coarsening
in Nimonic 115. Experimental data at short aging durations behaved in
accordance with the cubic rate law that corresponded to disappearance of the fine
followed by a plateau region where the number of primary precipitates was
nearly constant. At a later stage, data again followed the cubic rate law where
coarsening was deemed to have reached steady-state. The numerical model
predicted the trends in the experimental data reasonably well, better than the
existing unimodal LSW coarsening models [7]. The present work involves
experimental studies of coarsening in the temperature range of 800 through
1000C and subsequent treatment of bimodal coarsening kinetics using a
single consistent LSW rate equation.
Materials Characterisation V
253
2 Experimental procedure
The baseline material for coarsening studies was GTD111TM which was
solutionized at 1121C for 2h followed by an inert gas fan cool and subsequently
aged at 816C for 4h. In order to conduct careful and thorough coarsening
studies of GTD111TM, baseline specimens, 20 in number, were first examined to
obtain their key microstructural statistics, viz. secondary radius rS ' ,
secondary volume fraction f v _ S ' , tertiary radius rT '
and volume
3 Results
The baseline microstructure consisted of a bimodal distribution of cuboidal
secondary precipitates and much finer tertiary precipitates as shown in fig.
1(a). The corresponding bimodal size distribution (PSD) is shown in fig. 1(b).
The averaged baseline statistics from examination of 20 specimens are as
follows:
rS ' =266nm, f v _ S ' =37%, rT ' =34nm and f v _ T ' =4%.
The microstructures of specimens that are obtained upon aging will be
summarized next. At T1, the microstructure consisted of a bimodal distribution
after 250h of aging as seen from the micrograph in fig. 2(a). Upon further aging
to 4000h, the microstructures did not show any significant coarsening of the
secondary distribution (see fig. 2(b)); only a shape change from cuboids to
corner-rounded cuboids was noted after 2000h of aging. The number density of
tertiary precipitates decreased with progressive coarsening, also evident from the
micrographs. At T2, the tertiary distribution was only sighted at 200h (see fig.
2(a)), beyond which, the tertiary precipitates coarsened and merged with the
secondary PSD. Furthermore, shape change of the secondary precipitates from
WIT Transactions on Engineering Sciences, Vol 72, 2011 WIT Press
www.witpress.com, ISSN 1743-3533 (on-line)
Figure 1:
Materials Characterisation V
Figure 2:
255
55
fv_totbaselines
fv_totTHERMOCALC
fv (%)
50
45
40
35
30
T(C)
Figure 3:
7
6
fvT' (%)
5
4
3
2
1
0
0
1000
2000
3000
4000
5000
t(h)
Figure 4:
600
160
T1secondaryg'
T2secondaryg'
500
T4secondaryg'
120
100
300
80
rT' (nm)
T1tertiaryg'
400
rS' (nm)
140
T3secondaryg'
60
200
40
100
20
0
0
Figure 5:
1000
2000
3000
t(h)
4000
0
5000
Materials Characterisation V
257
with increasing temperature, also shown in fig. 3. The evolution of the tertiary
volume fraction upon aging at T1 is shown in fig. 4. The gradually decreasing
trend is expected for two reasons- (a) consumption of the tertiary by the
secondary during coarsening and (b) inherent coarsening of the tertiary , thus
merging with the secondary population.
Figure 5 shows a plot of the mean secondary radius versus aging time for
all the four data sets. The expected increase in rS ' due to coarsening is
preceded by a dip early on at T1, T2 and T3. The likely reason for this is
attributed to fraction of the coarser tertiary population joining into the secondary
distribution. The increase in the mean tertiary radius at T1 is also shown in the
same plot. However, the observed increase in rT ' is apparent because of the
fact that a fraction of the tertiary population is consumed during coarsening of
the secondary distribution (see fig. 4), which otherwise would have participated
in inherent coarsening of the tertiary distribution.
2.
3.
T2secondaryg'observed
T3secondaryg'observed
1.E+08
T4secondaryg'observed
rt3ro3 (nm3)
8.E+07
R=0.99
6.E+07
4.E+07
R=0.93
2.E+07
R=0.99
0.E+00
0
Figure 6:
500
1000
1500
t(h)
2000
2500
3000
Cubic rate law plot of the mean secondary radius at T2, T3 and
T4 (T2<T3<T4). Data used for analyzing kinetics correspond to
unimodal secondary PSDs only.
rSo '
k
is the radius at t 0 and k (T ) o e RT is the coarsening rate
time t,
T
constant with pre-exponential k o and activation energy Q . A plot of the cubic
rate law at T2, T3 and T4 shows reasonably good linear fits at all the three
temperatures as shown in fig. 6. As expected and congruent with microscopic
observations, the rate of coarsening (slope of the linear fit) is enhanced with
increasing aging temperature. Next, a plot of the coarsening rate
1
constant k (T ) versus
(see fig. 7) is used to extract the activation energy of
RT
coarsening as 212.45 (kJ/mol). The expression for the rate constant describing
the secondary coarsening kinetics is as follows:
rSo '
212447
k
k (T ) o e RT
T
(nm3/h)
(1)
Materials Characterisation V
259
18.5
k(T)(nm3/h)
18
Q=212,447kJ/mol
17.5
R=0.98
17
16.5
16
1/RT(mol/JK)
Figure 7:
1
Plot of k (T ) vs.
using the rate constants at T2, T3 and T4.
RT
The activation energy of secondary coarsening is obtained as
slope of the linear fit.
4.E+06
ObservedcoarseningrateTg'
ExpectedcoarseningrateTg'
3.E+06
PredictedcoarseningrateTg'
rt3ro3 (nm3)
3.E+06
R=1
2.E+06
R=0.98
2.E+06
1.E+06
5.E+05
R=0.80
0.E+00
0
Figure 8:
500
1000
1500
2000
t(h)
2500
3000
3500
4000
Cubic rate law plots of the mean tertiary radius at T1 shown for
three cases measured data, predicted dataset using secondary
kinetics and expected dataset by correcting for volume fraction loss
of tertiary during coarsening.
0.29 k o RT
k (T )
e
T
3
rTo '
(nm3/h)
(2)
and the corresponding linear fits for three cases (a) observed using measured
data, (b) predicted using eqn. (2) and (c) expected by correcting measured
data. The aging duration of 250h is taken to be the starting point i.e. t 0 and the
corresponding tertiary statistics are used to generate the predicted data set. The
expected tertiary radii are computed in accordance with assumption#3 by the
t
t
following procedure. Denote f v N_ T ' , N v N_ T ' and rTN'
to be the volume
fraction, particle density (number of precipitates per unit volume) and mean
radius of the tertiary distribution at the Nth aging time step corresponding to
time t t N . It should be noted that the data point at 250h is denoted using the
superscript t o ( t o 0 in this case) and longer aging times are scaled with
reference to 250h as the starting point. Now, f v N_ T1 ' f v N_ T ' is the tertiary
volume fraction which is consumed during secondary coarsening. This lost
fraction between successive time steps is accounted for to obtain the expected
results in a
with mean size rTtN' . The quantities VTtN' and V t N 1 t N are defined as
below:
t
VTN'
and
V
t N 1 t N
3
4
rTtN'
3
t N 1
tN
v _ T ' f v _ T '
t
N v N_ T '
(3)
(4)
Materials Characterisation V
261
N
rexp
ected _ T ' at t t N is computed as:
N
3
tN
t
t
VT ' V N 1 N
tN
1
(5)
rexp
ected _ T '
4
The correlation coefficient of the linear fit is improved from 0.8 for the
measured data to 0.98 (see fig. 8) for the expected data set by making this
correction. The agreement between the coarsening rate constants of the
expected and predicted datasets is also better when compared to the rate
constant of the measured dataset. The reasons for disparity between the rate
constants of the expected and predicted data sets could be two fold. First,
there could be errors associated with truncation of the tertiary PSD due to
overlap of the secondary and tertiary distributions, as is the case at 1000 and
2000h at T1. Secondly, any variations in the precipitate statistics that exist
among the baselines are ignored in the present analysis. It is quite possible that
the truncation errors and initial baseline variations could lead to errors in
estimation based on eqn. (5).
5 Conclusions
Aging experiments were conducted on GTD111TM in the temperature range of
800 to 1000C with the objective of studying bimodal coarsening kinetics in
this alloy. Secondary coarsening kinetics elucidated using data where the
PSDs were clearly unimodal yielded an activation energy of 212.45 (kJ/mol).
This kinetics was then applied to tertiary coarsening at the lowest aging
temperature by using an appropriate scaling factor to discount volume fraction
enhancement of coarsening. Separately, the kinetics of the measured data was
corrected approximately by accounting for volume fraction loss of tertiary
between successive aging steps accompanying secondary coarsening. The
coarsening rate constants obtained from these two approaches showed a
reasonable agreement with each other. Thus, the kinetics of bimodal
coarsening in GTD111TM could be described using a single consistent LSW rate
equation in the chosen range of temperatures.
Acknowledgements
The authors greatly appreciate S. Vipin for careful metallographic preparation of
GTD111TM aged specimens. Next, we would like to acknowledge the staff at the
National Metallurgical Laboratory, Jamshedpur, India for their assistance in
WIT Transactions on Engineering Sciences, Vol 72, 2011 WIT Press
www.witpress.com, ISSN 1743-3533 (on-line)
References
[1] Schilke, P. W. Advanced gas turbine materials and coatings. GE Energy.
Schenectady, NY : s.n.
[2] Sims, C. T. and Hagel, W. C. The Superalloys. New York : John Wiley &
Sons, 1972.
[3] Martin, J. W., Doherty, R. D. and Cantor, B. Stability of microstructure in
metallic systems. New York : Cambridge University Press, 1997.
[4] Dyson, B., Use of CDM in materials modelling and component creep lifeprediction. ASME Journal of Pressure Vessel Technology, 315,
pp. 281-296, 2000.
[5] Mastromatteo, F., Niccolai F., Giannozzi, M. and Bardi, U., The coarsening
kinetic of ' particles in nickel-based superalloys during aging at high
temperatures. Proceedings of the ASME Turbo Expo. 4, pp. 851-856, 2004.
[6] Stevens, R. A. and Flewitt, P. E. J., The effects of precipitate coarsening
during isothermal aging and creep of the nickel-base superalloy IN-738.
Materials Science and Engineering A 3, pp. 237-247, 1979.
[7] Coakley, J., Basoalto, H. and Dye, D., Coarsening of a multimodal nickelbase superalloy. Acta Materialia 58, pp. 4019-4028, 2010.
[8] Kelekanjeri, V. S. K. G., Moss, L. K., Gerhardt, R. A. and Ilavsky,
J., Quantification of the coarsening kinetics of precipitates in Waspaloy
microstructures with different prior homogenizing treatments. Acta
Materialia, 57(16), pp. 4658-4670, 2009.
[9] Sharghi-Moshatghin, R and Asgari, S., The effect of thermal exposure on
the characteristics in a Ni-base superalloy. Journal of Alloys and
Compounds 368, pp. 144-151, 2004.
[10] Chen, M. K. and Voorhees, P. W., The dynamics of transient Ostwald
ripening. Modelling and Simulation in Materials Science and Engineering,
1(5), pp. 591-612, 1993.
[11] Davies, C. K. L., Nash, P. and Stevens, R. N., The effect of volume fraction
of precipitate on ostwald ripening. Acta Metallurgica, 28(2),
pp. 179-189, 1980.
[12] Voorhees, P. W. and Glicksman, M. E., Solution to the multi-particle
diffusion problem with applications to ostwald ripening-II. Computer
simulations. Acta Metallurgica, 32(11), pp. 2013-2030, 1984.
Materials Characterisation V
263
Abstract
The reported studies are related to a new group of intelligent materials, such as
magnetorheological elastomers (MREs), which are composites of ferromagnetic
particles embedded in elastomer matrix. Studies on fabrication of MREs were
carried out using two different polyurethane elastomers as a matrix,
ferromagnetic particles and coupling agent. The matrices were differed in
elastomer stiffness and hardness. As a ferromagnetic component carbonyliron
powder with particle size of 6-9 m was used. Particles were oriented into chains
under the external magnetic field of 240 kA/m. Samples with anisotropic
particles arrangement and particles volume fraction equal to 11.5% were
examined. Microstructure of MREs was observed using Scanning Electron
Microscopy. Structural and magnetic anisotropy of the MREs was derived from
the magnetic studies. Rheological properties of the MREs, such as storage and
loss modulus and loss factor, were characterized as a function of shear frequency
and strength of the magnetic field. Absolute and relative magnetorheological
effects were calculated taking the microstructure into account, which was formed
in the course of the MRE fabrication, by changing the elastomer type and
application of the coupling agent. As a result of the studies it was found that the
MREs with stiffer matrix exhibit much lower MR effect than the soft ones. Also
the application of coupling agent lead to a decrease of MR effect by the
formation of bound elastomer surrounding the particles, which increases the
stiffness of the matrix.
Keywords: magnetorheological elastomers, polyurethanes, carbonyl iron,
rheological properties, coupling agents, microstructure, magnetic properties,
magnetorheological effect.
WIT Transactions on Engineering Sciences, Vol 72, 2011 WIT Press
www.witpress.com, ISSN 1743-3533 (on-line)
doi:10.2495/MC110231
1 Introduction
Magnetorheological elastomers (MREs) are composites of ferromagnetic
particles embedded in elastomer matrix. They belong to a new group of
intelligent materials, which change their rheological properties under the
influence of an applied magnetic field. They are solid analogues of the
magnetorheological fluids (MRF) where the matrix is a solid elastic polymer
rather than carrier oil. These both consist of micrometer-sized magnetically
permeable particles in a non-magnetic matrix material. As in the case of MRF,
the particles try to arrange themselves in the direction of the magnetic field. An
external magnetic field is applied during the curing process of the elastomer,
what is described in many papers, e.g. Zhou [1], Farshad and Benine [2], Jolly
et al. [3]. The field induces dipole moments within the particles, which seek
minimum energy states. Chains of particles with collinear dipole moments are
formed and curing of the polymeric matrix locks the chains in place. As a result
special structures, such as chains or columns of particles, remain in the matrix.
MREs have controllable magnetic field dependent modulus rather than field
dependent yield stress and also stable performance. The advantage of MREs, in
comparison to MRF, is that ferromagnetic particles do not undergo the
sedimentation. Moreover, the amount of filler is usually lower in MREs. As a
result, the weight of sensing and actuating devices based on MREs is reduced
and due to their microstructure, the time, following signal introduction, for a
strain change response to applying a magnetic field is shortened, as it was
described by Carlson and Jolly [4]. MREs exhibit reversible changes of their
properties and shape under the magnetic field, what makes them attractive for
applications as dampers, sensors or actuators, e.g. An and Shaw [5], Farshad and
Le Roux [6]. The automotive bushing based on MREs, adjusted to reduce
suspension deflection and to improve passenger comfort, were patented by
Watson [7] and Steward et al. [8] for Ford Motor Company.
The magnetic particles are usually carbonyl iron and as the matrix soft
silicone elastomers, poly(vinyl alcohol), gelatine, natural rubber or polyurethanes
can be used. The results of the studies presented in the literature [912] show
that the properties of MREs depend on the particles shape, size and volume
fraction, as well as the magnetic field strength. Only a few papers are focused on
the influence of the polymer matrix properties on the magnetorheological effect
(MR effect) in MREs. Gong et al. [13] and Lokander and Stenberg [9] reported
the advantageously influence on the MR effect of the addition of plasticizers to
the silicon rubber due to the decrease of the polymer matrix stiffness. Jiang et al.
[14] indicate on the positive effect of the particles surface modification on the
increase of the MR effect, while Zhang et al. [15], Fan et al. [16] and Li et al.
[17] show that the enhancement of the particle-polymer matrix interaction leads
to the decrease of MR effect due to the existence of bound-rubber phenomenon.
The improvement of filler-matrix interaction is an effective way to improve the
mechanical properties of materials; however in the case of MREs it can affect the
MR effect.
Materials Characterisation V
265
In our studies the influence of the matrix and the interaction between matrix
and filler on the MREs properties were investigated. We used two polyurethane
elastomers as a matrix, which were differed in stiffness and hardness. Particles
were oriented into chains under the external magnetic field of 240 kA/m.
Samples with the same particles volume fraction equal to 11.5%, were examined.
The iron particles-polyurethane matrix interactions were changed by the
application of silane coupling agent and the effect of the interaction on the
performance of MREs were investigated. Rheological properties of the MREs,
such as storage and loss modules and loss factor, were characterized as a
function of shear frequency and strength of the magnetic field. Absolute and
relative MR effects were calculated taking microstructure into account, which
was formed in the course the MREs fabrication, by changing the elastomer type
and application of the coupling agent.
Material
Viscosity
(mPas)
Density
(g/cm3)
PU 70/30
EPU 2.5
8000
1400
1.03
1.26
Glass transition
temperature (oC)
Soft
Hard
segments segments
-64
41
-23
165
Hardness
(oShA)
Youngs
modulus
(MPa)
< 10
87
0.1
14
Materials Characterisation V
267
a)
b)
Figure 1:
Although the viscosity and the chemical composition of the reactive mixtures
of substrates are different, the alignment of particles into chains in both cases
was obtained. The average distance among the chains of particles calculated
using image analysis, as described in Boczkowska et al. [19], is equal to 404
m.
3.2 Magnetic measurements
The existence of structural and magnetic anisotropy was confirmed by VSM
studies. Tests were carried out parallel (II) and perpendicular (L) to the direction
of particles chains, corresponding to the magnetic field direction during curing.
II
60
Magnetization [Am2/kg]
40
20
0
-800
-600
-400
-200
0
-20
200
400
600
800
-40
-60
-80
Magnetic field [kA/m]
b)
150
Magnetization [Am2/kg]
II
L
100
50
0
-800
-600
-400
-200
0
-50
200
400
600
800
-100
-150
Magnetic field [kA/m]
Figure 2:
In fig. 2 hysteresis loops obtained for the MREs with different matrix material
are shown. From the hysteresis loops the anisotropy coefficient (Ab) was
calculated at the selected value of magnetic field strength of 160 kA/m. It is
expressed by the ratio of magnetization measured at 160 kA/m, respectively
parallel and perpendicular to the particles alignment direction. The Ab values are
similar for both kind of MREs matrix. They are equal to 1.50 for PU 70/30 and
1.42 for EPU 2.5, what means that the difference in viscosity and the chemical
WIT Transactions on Engineering Sciences, Vol 72, 2011 WIT Press
www.witpress.com, ISSN 1743-3533 (on-line)
Materials Characterisation V
269
0.90
0.80
GI [MPa]
0.70
11,5%Fe 80 kA/m
11,5%Fe 320 kA/m
11,5%Fe+1%U-15 0 kA/m
11,5%Fe+1%U-15 160 kA/m
11,5%Fe+1%U-15 480 kA/m
0.60
0.50
0.40
0.30
0.20
0.10
0.00
0.1
0.16
f [Hz]
10
100
f [Hz]
10
100
11,5%Fe 0 kA/m
11,5%Fe 80 kA/m
11,5%Fe 160 kA/m
11,5%Fe 320 kA/m
11,5%Fe 480 kA/m
11,5%Fe+1%U-15 0 kA/m
11,5%Fe+1%U-15 80 kA/m
11,5%Fe+1%U-15 160 kA/m
0.14
0.12
GII [MPa]
0.10
0.08
0.06
0.04
0.02
0.00
0.1
0.35
0.3
tan
0.25
0.2
0.15
0.1
11,5%Fe 0 kA/m
11,5%Fe 160 kA/m
11,5%Fe 480 kA/m
11,5%Fe+1%U-15 80 kA/m
11,5%Fe+1%U-15 320 kA/m
0.05
0
0.1
Figure 3:
f [Hz]
11,5%Fe 80 kA/m
11,5%Fe 320 kA/m
11,5%Fe+1%U-15 0 kA/m
11,5%Fe+1%U-15 160 kA/m
11,5%Fe+1%U-15 480 kA/m
10
100
Changes of storage (G), loss (G) modulus and loss angle (tan )
as a function of frequency under magnetic field of 0, 80, 160, 320,
480 kA/m.
Materials Characterisation V
271
0.70
0%U-15
0.60
1%U-15
G' [MPa]
0.50
0.40
0.30
0.20
0.10
0.00
0
80
160
240
320
400
480
400
480
560
H [kA/m]
800
0%U-15
700
1%U-15
G'/G'0 [%]
600
500
400
300
200
100
0
0
Figure 4:
80
160
240
320
H [kA/m]
560
Both MR effects, absolute and relative, grow with the magnetic field growth.
The significant difference is observed between samples fabricated with and
without application of silane coupling agent. The difference is much more
significant with the magnetic field growth. The MR effect is higher in the MREs
obtained without the application of the coupling agent because of the lower
stiffness of the polyurethane matrix, what makes the ferromagnetic particles
easier to attract under the magnetic field and as a result to introduce the
additional strain and stress to the polyurethane matrix.
Similar results were obtained for MREs fabricated from elastomers with
different stiffness and hardness, as shown in fig. 6.
WIT Transactions on Engineering Sciences, Vol 72, 2011 WIT Press
www.witpress.com, ISSN 1743-3533 (on-line)
0.30
PU 70/30
0.25
EPU 2,5
G' [MPa]
0.20
0.15
0.10
0.05
0.00
0
80
160
240
320
400
480
560
400
480
560
H [kA/m]
600
PU 70/30
500
EPU 2,5
G'/G'0 [%]
400
300
200
100
0
0
80
160
240
320
H [kA/m]
Figure 5:
MREs with the stiffer matrix (EPU 2.5) exhibit significant lower MR effect
because the stiffer matrix makes impossible the particles to displace when they
are subjected to the magnetic field.
4 Conclusions
The properties of the MREs elastomeric matrix have a great influence on their
rheological properties. Elastomer with higher stiffness and hardness leads to the
higher storage and loss modulus of the MREs measured without magnetic field,
while under the magnetic field the values are significant lower in comparison to
Materials Characterisation V
273
MREs obtained from soft elastomer. As a result the absolute and relative MR
effect is much higher in MREs obtained from soft elastomer.
The highest absolute and relative MR effect was obtained for the MREs
synthesized from PU 70/30 elastomer with 11.5 vol.% of carbonyl-iron particles.
The values of MR effects under the magnetic field of 480 kA/m reached 0.27
MPa and 540%, respectively, when the particles chains were aligned parallel to
the magnetic field direction or 0.52 MPa and 750%, when the chains of particles
were sloped to 45 degrees. The influence of the particles chains arrangement on
the rheological properties is the subject of our different study.
The increasing of the adhesion between particles and polyurethane matrix
leads to the significant decrease in the MR effect by the existence of bound
elastomer around the particles, what makes the matrix stiffer. High stiffness of
the MREs matrix do not allow for the particles displacement under the magnetic
field and subsequently the deformation of the elastomeric matrix.
The rheological properties of MREs depend also on the oscillation frequency
and magnetic field strength.
Acknowledgement
This work was financed by National Centre for Research and Development
(Poland) as a grant no. NR 15 0010 04.
References
[1] Zhou G.Y., Shear properties of magnetorheological elastomer, Smart
Materials and Structures, 12, pp. 139-146, 2003.
[2] Farshad M. and Benine A., Magnetoactive elastomer composites, Polymer
Testing, 23, pp. 347-353, 2004.
[3] Jolly M. R., Carlson J.D., Munoz B.C. and Bullions T.A., The
Magnetoviscoelastic Response of Elastomer Composite Consisting of
Ferrous Particles Embedded in a Polymer Matrix, Journal of Intelligent
Material Systems and Structures, 7, pp. 613-622, 1996.
[4] Carlson J. D., Jolly M. R., MR fluid, foam and elastomer devices,
Mechatronics, 10, pp. 555-569, 2000.
[5] An Y., Shaw M.T., Actuating properties of soft gels with ordered iron
particles: basis for a shear actuator, Smart Materials and Structures, 12,
pp.157-163, 2003.
[6] Farshad M., Le Roux M., Compression properties of magnetostrictive
polymer composite gels, Polymer Testing, 24, pp.163-168, 2005.
[7] Watson J. R., Methods and apparatus for varying the stiffness of a
suspension bushing, US Patent 5.609.353, 1997.
[8] Stewart W. M., Ginder J. M., Elie L. D., Nicholls M. E., Method and
apparatus for reducing brake shudder, US Patent 5.816.587, 1998.
[9] Lokander M., Stenberg B., Improving the magnetorheological effect in
isotropic magnetorheological rubber materials, Polymer Testing, 22, pp.
677-680, 2003.
WIT Transactions on Engineering Sciences, Vol 72, 2011 WIT Press
www.witpress.com, ISSN 1743-3533 (on-line)
Materials Characterisation V
275
Abstract
In this work we include valuable information about the way in which nickel may
become a valuable smart material for some applications and we provide a better
understanding of the influence of internal stresses on its magnetoelastic effects.
The different states of internal stress are achieved via different heat treatments
obtained by modifying three main parameters: the heating temperature, the heating
time and the cooling method. Then, we carried out the next works: first, a
microscopic analysis of the grain size of the samples under different heat
treatments; second, all the tested specimens were subjected to an exhaustive
internal stress analysis via X-Ray diffraction techniques; and third, the E- and
-effects were estimated in order to link the internal stress state of the sample to
its magnetoelastic response. The results can guide us in selecting the most suitable
heat treatment in order to make nickel show the smart properties we desire.
Keywords: magnetoelasticity, internal stress, heat treatments, nickel, X-Ray
diffraction.
1 Introduction
Magnetomechanical materials are a kind of smart material in which there is a
reciprocal coupling between their mechanical and magnetic properties [1]. In this
broad category, described by Jiles in [2], we can find both magnetoelastic and
magnetoplastic materials. The former are those in which the variation of the
altered properties is reversible, whereas the latter are those in which the recovery
of the properties is not necessarily obtained simply on removal of the magnetic
or mechanical action.
WIT Transactions on Engineering Sciences, Vol 72, 2011 WIT Press
www.witpress.com, ISSN 1743-3533 (on-line)
doi:10.2495/MC110241
2 Experimental details
Pure crystalline rods of nickel 201, all from the same molten material and
manufacturing process, were used in this work. They were 110 mm in length and
10 mm in diameter and they showed a purity level of 99.90% and a density of
8912 kg/m3.
The magnetoelastic properties were measured via a recently experimental
system which we developed in [9] and enhanced later in [3]. It is based on laser
Doppler vibrometry and it turns out to be a suitable method for measuring Eand - effects: not only is it able to measure both effects simultaneously and
accurately, but it also allows us to obtain stress- and path-dependencies.
The different heat treatments are achieved by modifying three main
parameters: the heating temperature, the heating time and the cooling method.
The selection of the different heating temperatures for our scheduled heat
treatments, probably the more significant parameter from the point of view of
internal stresses, was made in accordance with the recommendations found in
[10] for nickel specimens. This led us to three different treatments: annealing
(7051205C), designed to produce a recrystallized grain structure and softening
in work-hardened alloys; stress relieving (425870C), used to reduce stresses in
work-hardened non-age-hardenable alloys without recrystallizing the grain
structure; and stress equalizing (230315C), used to balance stresses in coldworked materials without appreciably decreasing mechanical strength. Table 1
presents a summary of the heat treatments. In it, each specimen has been
identified by three digits (which represent the heating temperature in Celsius),
two digits (which indicate the heating time in hours) and one letter (which refers
Materials Characterisation V
277
to the cooling method, i.e., water, air or furnace). The initial state of the nickel
specimens is labelled AR (as received).
Finally we want to emphasize that, leaving aside field-dependence, other
influences such as stress-dependence and path-dependence must be considered in
our measurements [3]. In order to avoid their influences, we will report only
those results which were obtained under 0.50 MPa and following a
magnetization path from the state of zero retentivity (Mr0) to the state of positive
saturation magnetization (Ms+).
Table 1:
Specimen
0
1
2
3
4
5
6
7
8
9
10
11
12
13
Heating
temp. (C)
300
300
600
600
900
900
900
900
900
900
900
900
900
Cooling
method
Water
Air
Water
Air
Water
Air
Furnace
Water
Air
Furnace
Water
Air
Furnace
Nomenclature
AR
300-02-W
300-02-A
600-02-W
600-02-A
900-02-W
900-02-A
900-02-F
900-04-W
900-04-A
900-04-F
900-08-W
900-08-A
900-08-F
Figure 1:
Materials Characterisation V
279
On the one hand, specimens which were subjected to stress relieving (300-02W and 300-02-A) and stress equalizing (600-02-W and 600-02-A) processes
show a grain size similar to that of specimen AR. This fact agrees with the fact
that stress relieving and stress equalizing reduce and balance stresses without
recrystallizing the grain structure. Besides, in spite of small variations, we can
also observe higher grains in specimens which were subjected to stress
equalizing due to the higher heating temperature and consequently the higher
grain boundary diffusion. On the other hand, annealed specimens (900-02-W,
900-02-F and 900-08-W) clearly shows much larger grains, which will translate
into lower internal stresses. Specimen 900-08-F has not been shown because it
has got almost the same distribution of grains due to the long heating time, which
means high grain boundary diffusion even when cooled in water. Furthermore,
the concentration of small grains which appears in specimen 900-02-W
corresponds to the centre of the specimen, where the heating time was not high
enough to continue increasing the size of inner grains (the diffusion processes
goes from the outside to the inside of the rod).
Table 2:
Specimen
AR
300-02-W
300-02-A
600-02-W
600-02-A
900-02-W
900-02-A
900-02-F
900-04-W
900-04-A
900-04-F
900-08-W
900-08-A
900-08-F
Microscopic measurements.
Small grains
Perim.
Area
(m)
(m2)
350
8200
440
11900
520
14400
600
16900
700
21500
530
16600
620
20700
620
23600
-
Medium grains
Perim.
Area
(m)
(m2)
1240
41000
1240
53000
1720
144000
1730
157000
1660
147000
1640
137000
1820
167000
1750
143000
1960
194000
1900
195000
1870
170000
Large grains
Perim.
Area
(m)
(m2)
4000
590000
4200
510000
4100
650000
4200
670000
4400
670000
4700
660000
5300
780000
5300
720000
5600
940000
Table 2 shows the measurements of grain perimeter and grain area for each
specimen. The results seemed to indicate the existence of three different
distributions: small grains, with a perimeter lower than 1000 mm; medium
grains, with a perimeter between 1000 and 3000 mm; and large grains, with a
perimeter greater than 3000 mm. Specimens AR, 300-02-W and 300-02-A only
contain small grains, although perimeters and areas increase slightly when the
specimens have been heated. Specimens 600-02-W and 600-02-A show small
grains mainly although medium grains are not uncommon. Finally, all specimens
heated to 900C show large distributions of medium and large grains due to
WIT Transactions on Engineering Sciences, Vol 72, 2011 WIT Press
www.witpress.com, ISSN 1743-3533 (on-line)
Materials Characterisation V
281
Figure 2:
Figure 3:
Detail of the second peak of the X-Ray diffraction profile for the
specimen 900-02-W and the fitting procedure in WinPLOTR.
Now, considering that we can make the correction for the instrumental
broadening, the corrected integral breath due to microstrain and particle size can
be written as follows:
.
(2)
Figure 4:
where the first addend is the particle size contribution and the second addend is
the strain () contribution. Then, left-multiplying by cos we obtain:
cos
4 sin
(3)
Materials Characterisation V
283
Regarding the lattice parameter, no significant variations were found, but the
obtained value is in good agreement with the typical values of the lattice
parameter in nickel samples (around 3.5 ).
Table 3: Internal stresses and lattice parameter.
Specimen
AR
300-02-W
300-02-A
600-02-W
600-02-A
900-02-W
900-02-A
900-02-F
900-04-W
900-04-A
900-04-F
900-08-W
900-08-A
900-08-F
Internal
stress (MPa)
479.0
579.6
536.6
475.2
498.3
469.0
276.9
223.4
356.6
248.6
106.8
415.8
207.6
86.2
Microstrain
(%)
0.23
0.27
0.25
0.22
0.24
0.22
0.14
0.11
0.17
0.13
0.05
0.20
0.10
0.04
Elastic
modulus (GPa)
213
213
213
216
210
220
201
199
213
203
203
216
208
203
Lattice
parameter ()
3.53
3.53
3.53
3.53
3.53
3.53
3.54
3.53
3.54
3.54
3.54
3.52
3.53
3.53
Figure 5:
Figure 6:
Since it is not the main goal of this work, we will not include results about the
E- and - effects of each treated specimen. The interested reader may consult
such results in our work [4].
Regarding the influence of the parameters involved in the heat treatments, we
can claim than the heating temperature is the most significant parameter since
only temperatures higher than 700C (900 in our tests) may lead to a pattern II
behaviour. The cooling methods also stand out as a significant parameter since
water cooling may avoid the grain diffusion and increase internal stresses, what
would lead to a pattern I behaviour. Finally, the heating time has a very minute
influence on the magnetoelastic behaviour when using long cooling methods
such as air cooling or furnace cooling. This can be seen in Fig. 6.
Another important feature we should analyze is the influence of the heat
treatments and the internal stresses on the magnetic field from which the elastic
modulus or the magnetomechanical damping remains the same, i.e., on the
magnetic field which makes the E- and - effects saturate. We can conclude
that in all the cases the saturating magnetic field is around 350 Oe, so it does not
depend on any of the heat treatment parameters which were considered.
WIT Transactions on Engineering Sciences, Vol 72, 2011 WIT Press
www.witpress.com, ISSN 1743-3533 (on-line)
Materials Characterisation V
285
4 Conclusions
The previous results can guide us in selecting the most suitable heat treatment in
order to make nickel show the smart properties we desire. It is important to
notice that the optimization will depend on the objective function we select.
Smart materials are usually required to provide a maximum variation of the
salient magnitude by supplying a minimum modification of the input. This fact,
from the point of view of a magnetoelastic material, means large variations in
applied forces, natural frequencies and damping ratios obtained via small
changes in the applied magnetic field. Nevertheless, we should not discard that,
in some cases, the objective is the opposite: to obtain a constant value of elastic
modulus and damping along a range of magnetic field as large as possible.
We can extract valuable information about the different possibilities of
optimization from Fig. 5. On the one hand, if nickel is required to act as a smart
material which provides the highest E- and - effects, then it should behave
following pattern II. In addition, the characteristic damping peak of pattern I
disappears, so we get a more predictable behaviour. Thus, in the light of section
3 results, the material should show very low internal stresses, which can be
achieved via annealing heat treatments with slow cooling velocities (900-0X-A
or 900-0X-F). On the other hand, one may need that nickel show elastic and
damping properties as constant as possible under different applied magnetic
fields. Although this behaviour cannot be perfectly fitted, it is possible to
minimize the E- and - effects by making the material follow pattern I.
Again, we know from discussion in section 3, that this means getting internal
stresses as high as possible, which can be achieved via work-hardening, stress
equalizing heat treatments and fast cooling velocities (AR, 300-0X-W).
Leaving aside optimization rules, this work also provides relevant results,
details and discussions related to the microscopic and internal stress
characterization of this material. We have carried out an in-depth material
characterization in terms of grain distribution and internal stresses when nickel
specimens are subjected to several different heat treatments. On the one hand we
have included many micrographies which help us to understand not only the way
does each heat treatment modify the grain size, but also how the grain diffusion
acts depending on the heating temperature, the heating time and the cooling
method. On the other hand we have carefully described the way we have
obtained the microstrain, internal stress and lattice parameter in all our
specimens via X-Ray diffraction and the Williamson-Hall method. All these
results agree with the fact that the internal stress decreases when we increase the
heating temperature and decrease the cooling velocity, the heating time
becoming an almost negligible parameter.
References
[1] du Trmolet de Lacheisserie, E., Magnetostriction: Theory and
Applications of Magnetoelasticity, CRC Press: Boca Raton, 1993.
Materials Characterisation V
287
Abstract
Low water-to-cement ratio (w/c) reactive powder concretes (RPCs) exhibit much
higher compressive strengths compared to conventional concrete through
optimized particle packing and specialized curing regimes. The high strain-rate
impact behavior of RPCs was investigated at the macroscale. However, little
work has been done to study the fundamental material behaviors and failure
mechanisms of RPC under high strain impact and penetration loads at lower
length scales. These high strain-rate loadings have many possible effects on
RPCs at the microscale and nanoscale, including alterations in the composition
and bonding present in hydrated phases such as calcium silicate hydrate (C-S-H),
in addition to fracture and debonding. In this work, the possible chemical and
physical changes of RPCs under high strain-rate impact and penetration loads
were investigated using a novel technique wherein nanoindentation
measurements were spatially correlated with chemical composition using
electron microscopy. Results indicate that high strain-rate impacts degrade both
the elastic modulus and indentation hardness of RPCs and, in particular C-S-H,
with damage likely occurring due to microfracturing and debonding. Additional
studies will be required to better understand degradation phenomena within
C-S-H itself.
Keywords: nanoindentation, SEM, EDX, RPC, UHPC, C-S-H.
1 Introduction
Concretes and Portland cement-based materials are the most-produced manmade materials on earth, with over twenty billion tons produced per year. The
WIT Transactions on Engineering Sciences, Vol 72, 2011 WIT Press
www.witpress.com, ISSN 1743-3533 (on-line)
doi:10.2495/MC110251
Materials Characterisation V
289
2 Experimental procedures
2.1 Materials
An RPC mixture with a w/c of 0.2 and an unconfined compressive strength of 180
MPa was cast into panels 305-mm wide by 305-mm long and 25.4-mm thick for
ballistic impact testing. The panels were impacted by 11-mm-diameter steel
spheres at half the ballistic limit velocity, which was determined according to
MIL-STD-662 [11]. The mixing, casting, curing, and ballistic testing is detailed
in the manuscript by Rushing et al. [2].
2.2 Coupled nanoindentation and SEM/EDX analysis
Analysis of impacted and non-impacted RPC specimens was performed using a
novel technique of nanoindentation coupled with SEM imaging and EDX
chemical analysis. In this coupled method, a large number of indents are
performed over a standardized indentation grid placed on the RPC sample, after
which, each indent is spatially correlated using SEM to obtain an image and
EDX to determine the chemical composition. The resulting dataset contains the
nanomechanical properties and chemical composition along with an image at
each indentation site. This coupled method allows for improved differentiation
between the various components present in RPCs and can be used to better
correlate alterations in nanomechanical (e.g., due to impact loadings) properties
to specific microstructural features. Details on specimen preparation techniques
and experimental methods are described below.
2.2.1 Specimen preparation
Specimens were extracted from the panel in the impacted zone and a nonimpacted zone as shown in Figure 1. The center of the non-impacted location
was 38 mm from both the top edge and side of the panel. This location was
selected to minimize the potential for edge effects from casting while also
avoiding the damaged zone to the greatest extent. A 25.4-mm-diameter diamondtipped coring bit was used to core the non-impacted specimens, which were then
cast into 31.8-mm-diameter cylindrical molds using EpoHeat low-viscosity
epoxy supplied by Buehler. The impacted specimen was sectioned using an oilcooled Struers Secotom high precision cut-off saw. The cross section of the
specimen was placed into a 31.8-mm-diameter cylindrical mold and mounted in
EpoHeat epoxy from Buehler. After the epoxy fully cured, the samples were
sectioned in half by the oil-cooled cut-off saw to obtain a cross section from the
center of the panel, thus limiting any surface effects such as laitance.
Nonimpacted
specimen
location
Line on
crater
represents
edge of
crosssection
analyzed
Impacted
specimen
location
Figure 1:
Materials Characterisation V
291
LDCSH
HDCSH
Aggregate
Cement
AberrantTest
2.00
1.75
Load(mN)
1.50
1.25
1.00
0.75
0.50
0.25
0.00
0
25
50
75
100
125
150
175
200
225
250
Depth(nm)
Figure 2:
Materials Characterisation V
293
18
Frequency(%)
16
14
12
10
8
6
4
2
200
190
180
170
160
150
140
130
120
110
100
90
80
70
60
50
40
30
20
10
ElasticModulus(GPa)
Figure 4:
Materials Characterisation V
295
16
47.1%
14
NonImpacted
Impacted
Frequency(%)
12
10
8
6
4
2
15
14
13
12
11
10
IndentationHardness(GPa)
Figure 5:
Frequency(%)
25
Impacted
20
15
10
5
50
45
40
35
30
25
20
15
ElasticModulus(GPa)
Figure 6:
Frequency(%)
40
35
NonImpacted
30
Impacted
25
20
15
10
5
2.25
2.00
1.75
1.50
1.25
1.00
0.75
0.50
0.25
IndentationHardness(GPa)
Figure 7:
Materials Characterisation V
297
4 Conclusions
A novel statistical nanoindentation technique was developed that spatially
correlated the location of indents performed with an image and chemical
analyses using EDX. Using this method, the influence of high strain-rate impact
loadings on the nanomechanical properties of RPCs was determined. Significant
degradation in both the elastic modulus and hardness occurred due to impact
loadings of RPCs. Closer examination of C-S-H in particular also showed
degradation. However, a link between chemical changes in the C-S-H and
deterioration in nanomechanical properties could not be made. Based on analysis
of all indents performed on RPC specimens, it is likely that a much of the
degradation in elastic modulus and hardness observed stems from
microfracturing and debonding, which occurs due to the impact.
Future work involves additional experimental studies and quantitative data
analysis to further evaluate the possible mechanisms of degradation present in
RPCs.
Acknowledgement
Permission to publish this article was granted by Director, Geotechnical &
Structures Laboratory.
References
[1] Millard, S.G., et al., Dynamic enhancement of blast-resistant ultra high
performance fibre-reinforced concrete under flexural and shear loading.
International Journal of Impact Engineering, 2010. 37: p. 405-413.
[2] Rushing, T.S., et al., Independent effects of matrix strength and fiber
reinforcement on concrete's ballistic resistance. SAVIAC.
[3] Constantinides, G. and F.-J. Ulm, The nanogranular nature of C-S-H.
Journal of the Mechanics and Physics of Solids, 2007. 55: p. 65-90.
[4] Grady, D., Shock equation of state properties of concrete, in Proc. of
Structures under Shock and Impact IV, N. Jones et al., Editor. 1996,
Computational Mechanics Publications: Southampton, UK. p. 405-414.
[5] Velez, K., et al., Determination by nanoindentation of elastic modulus and
hardness of pure constituents of Portland cement clinker. Cement and
Concrete Research, 2001. 31: p. 555-561.
[6] Hughes, J.J. and P. Trtik, Micro-mechanical properties of cement paste
measured by depth-sensing nanoindentation: a preliminary correlation of
physical properties with phase type. Materials Characterization, 2004. 53:
p. 223-231.
[7] Ulm, F.-J., et al., Statistical indentation techniques for hydrated
nanocomposites: concrete, bone, and shale. Journal of the American
Ceramic Society, 2007. 90(9): p. 2677-2692.
Materials Characterisation V
299
Abstract
The dynamic characteristics of concrete are important for structures subjected to
earthquake actions. Now there is much experimental data about this, and most
are uniaxial test data. In large concrete structures, such as concrete dams,
concrete reactive power, the stress states usually are in multiaxial states;
therefore the main objective of this paper is to study the multiaxial compressive
characteristics of concrete subjected to high strain rates under multiaxial
compressive loading. Cubic specimens (100mm by side) are subjected to quasistatic and dynamic proportional biaxial and triaxial compression tests. The strain
rates used are 10-5/s , 10-4/s, 10-3/s, 10-2/s. The stress ratios are for the biaxial
compression 1:0.1, 1:0.25, 1:0.5, 1:0.75,1:1; for the triaxial compression the
applied constant confining pressures are 4MPa, 8MPa, 12MPA, 16MPa. The
tests are carried out on a concrete triaxial dynamic test system designed by the
authors. The strength characteristics of the specimens at different strain rates and
stress ratios are given. The experiment results indicate that the dynamic strength
of concrete under multiaxial compressive stress states is higher than that under
the uniaxial compressive stress state. Based on the test data, the failure criterion
is established on the octahedral stress space. Its characteristic is that the effect of
the similar angle on the dynamic strength is considered to reflect the changes of
the dynamic strength between the tensile and compressive meridians.
Keywords: dynamic strength, strain rate, biaxial stress state, triaxial stress state.
1 Introduction
Dynamic loading on concrete structures arising from natural hazards such as
tornadoes, earthquakes and ocean waves is of great practical significance. Under
such dynamic conditions, the loading-rate dependence of material response
causes the material behaviour to be significantly different from what is observed
under quasi-static conditions. Hence a thorough knowledge of material
constitutive relationships and failure criterion, which cover a wide range of strain
rates, is very important for the design of structures subjected to all types of
loading likely to be encountered during the design lifetime.
Bischoff and Perry [1] reviewed and analyzed the response of concrete under
dynamic loads and discussed factors that influence the dynamic compressive
behaviour of concrete, such as concrete quality, aggregate type, age, curing, and
moisture conditions. Malvar and Ross [2] reviewed the existing data describing
the effects of strain rate on the compressive strength of concrete and compared
the dynamic increase factor (DIF) formulation recommended by the European
CEB Model Code [3]. However; extremely rare dynamic experiments in
multiaxial stress states are available among the current documents. For material
in multiaxial stress states, cases with biaxial compressive or triaxial compressive
loading with the other side keeps constant confining pressure are the two typical
loading patterns, which is of great significance for studying the behaviour under
arbitrary multiaxial stress states. To improve understanding of the mechanical
behaviour of concrete, experiments on the dynamic behaviour of concrete under
biaxial and triaxial dynamic compression stress states were conducted in this
research corresponding to the range of strain rates and stress states encountered
in engineering practice.
2 Experimental program
2.1 Preparation of specimens
Plain concrete cubes with a size of 100100 100 mm were subjected to biaxial
or triaxial dynamic compressive loads. Commercially available Portland cement
was used. Crushed natural stones were used as coarse aggregate with maximum
particle size of 20 mm. River sand was used as the fine aggregates. The concrete
mixture proportions by weight are water: cement: fine aggregate: coarse
aggregate =1.00:1.02:4.38:5.35. All specimens required for the aforementioned
tests were cast on the same day for each mix and covered with a plastic sheet to
prevent moisture loss. They were demoulded after 1 day and cured in the fog
room at a relative humidity of 95% and temperature of 272C till the age of 7
days. The 28-day compressive strength of concrete obtained by testing standard
cube specimens (150mm150mm150mm) is 20 MPa. Then, the specimens are
dried in air for 8 weeks before testing.
Materials Characterisation V
301
Figure 1:
Figure 2:
The testing system allowed free and independent motion in three directions.
Along each direction, a pair of pressure levers loaded a test specimen through
two platens located on both sides of the specimen. A spherical hinge was
installed between a lever and a platen on the same side of the specimen to ensure
WIT Transactions on Engineering Sciences, Vol 72, 2011 WIT Press
www.witpress.com, ISSN 1743-3533 (on-line)
0
9.84
10.63
11.38
12.32
Strain
rate:/s
10-5
10-4
10-3
0
9.84
10.63
11.38
4
30.05
32.11
33.70
1
14.00
15.32
16.66
18.01
16
72.14
75.34
74.08
Materials Characterisation V
303
2
2
fus
1 1
(1)
where s is the quasi-static strain rate, its value being selected as 10-5/s in this
.
paper; is the current strain rate; fus is the uniaxial compressive strength of
concrete at quasi-static loading; fbd is the dynamic strength of concrete in biaxial
stress state; =2/1 is the stress ratio; P1, P2, P3 and P4 represent the parameters
associated with material properties. By fitting to the test, data, P1, P2, P3 and P4
are determined as 0.446, 0.0875, 1.43 and 6.42 respectively. The multiple
correlation coefficient being 0.9580, and the mean error being 0.4122 MPa. The
suggested relationship in eqn (1) is depicted in Figure 3 and the test results are
also shown for comparison. Fairly good agreement is achieved.
.
Figure 3:
f
A( ) lat B() lat 1 c 0
f
f
f
us
us
us
(2)
(a)
(b)
(c)
Figure 4:
Materials Characterisation V
305
oct
f cs
a1 b1 oct c1 oct
f cs
f cs
(3)
;a1,b1,c1
are parameters depending on the loading rate, which can be determined by fitting
to the test data, as listed in Table 3.
Table 3:
Strain rate/s-1
10-5
10-4
10-3
10-2
a1
0.2033
0.2865
0.3550
0.4187
Fitting results.
b1
0.9730
0.7384
0.5723
0.4708
c1
0.4816
0.2948
0.1647
0.0975
R2
0.9547
0.9497
0.9505
0.9407
Figure 5:
5 Conclusions
The dynamic behavior of concrete in biaxial and triaxial stress state has been
experimentally studied under linearly increasing loads of high strain rates. Based
on the extensive test data and analyses, the following conclusions can be drawn:
(1) The ultimate strength of plain concrete nonlinearly increases with its
confining pressure at all load/strain rates that were considered in this study, but
the magnitude of increment depends on the lateral stress ratio.
(2) At low confining pressure, the ultimate strength of concrete increased
with the strain rate. When the confining pressure was approximately higher than
the uniaxial static strength, the ultimate strength tended to vanish with the strain
rate.
(3) Under biaxial and triaxial stress state, the failure envelope in the
octahedral stress space gradually expands with the strain rate.
(4) The proposed united strength criterion for concrete under multiaxial
stress state reasonably reflects both the effect of strain rate and the effect of
lateral confinement. Fairly good agreement with experimental results is
achieved.
Acknowledgements
This study was supported by the National Natural Science Foundation of China
under Grants 90815026 and 51079019 at Dalian University of Technology,
Dalian, China, and Grant 50908026 at Dalian University, Dalian, China.
References
[1] Bischoff, P. H., Perry, S. H., Compressive Behavior of Concrete at High
Strain Rates. Materials and Structures, 24 (2), pp. 425-450, 1991.
[2] Malvar, L. J., Ross, C. A., Review of Strain Rate Effects for Concrete in
Tension. ACI Materials Journal, 95(6), pp. 435-439, 1998.
[3] Comit Euro-International du Bton, Model Code 90, CEB-FIP, Redwood
Books: Trowbridge, Wiltshire, UK, pp. 48-51, 1990.
[4] Newman, J. B., Concrete under Complex Stresses. Development in Concrete
Technology-1, F. D. Lydon, ed., Applied Science Pub: London, UK, pp.
151-219, 1979.
Materials Characterisation V
307
Abstract
This work deals with the simulation of the rutting resistance of bituminous
binders used in road pavement. First, the experimental protocol was assessed to
simulate pavement traffic allowing the prediction of the rutting depth evolution
versus cycle number of wheel passes under isothermal conditions. Then a
probabilistic parametric approach was developed to take into account the
different parameter uncertainties related to the changes in experimental
conditions. We investigate through the stochastic approach the rutting sensitivity
of bituminous mixes under traffic load. A confidence region of a high probability
of 99% is defined to allow the prediction of the in situ rutting potential of
bituminous specimens.
Keywords: rutting, bitumen, uncertainties, stochastic.
1 Introduction
The environment protection has been and continues to be the major concern of
the road politics in industrialized countries. With the signature of the Kyoto
protocol agreement, the European Union is involved to reduce the energy
consumption and the resulting greenhouse gases emissions [1]. The warm mix
asphalt (WMA) is a promising technology which has been developed to
contribute to the protection of the environment and sustainable development
program [25]. The WMA is a mixture of mineral aggregates and bitumen
produced and placed on the road at a lower temperature comparing to the
traditional hot mix asphalt (HMA). Reductions in temperature of 20-50C have
WIT Transactions on Engineering Sciences, Vol 72, 2011 WIT Press
www.witpress.com, ISSN 1743-3533 (on-line)
doi:10.2495/MC110271
0/2 Noubleau
2/6 Noubleau
6/10 Noubleau
33%
20%
45%
Filler calcaire
Piketty
2%
35/50
Bitumen
5%
Materials Characterisation V
309
binder so that it can be sufficiently viscous to coat the aggregate, a surface active
additive was incorporated at 0.3% into the bitumen.
2.2 Laboratory compaction
To produce representative specimens for rutting test, slabs were manufactured
with the prepared bituminous mixes and afterwards compacted in moulds by a
roller compactor device (LCPC France) according to the standard
specifications (NF EN 12679-33). The apparatus (Fig. 1) provides a
pneumatically powered means of compacting slabs to reproduce in situ
compaction. The dimensions of the obtained slabs are 500 mm x 180 mm x
100 mm.
Figure 1:
Figure 2:
100
(1)
where r0i represents the initial measurements at the positions (i) located on the
slab and h its thickness.
In figure 3, we present the evolution of the rut depth in percent versus the
number of cycles for both hot and warm asphalt mixes. One can see that the 2
mixtures exhibit nearly the similar performance with a better rutting resistance of
the HMA particularly with the increasing of the cycles number. In fact, we
recorded after 30 000 cycles a rut depth of 4.23% for the HMA versus 6.53% in
the WMA case. We note that in the French specifications [8], the maximum of
rut depth shall be less than 10%. This constraint is marked by a dark continuous
line in figure 3.
Rut(%)
10
HMA
WMA
0.1
100
1000
10000
Cycles
Figure 3:
100000
Materials Characterisation V
311
S ( X ) p X ( x ) log( p X ( x )) dx
(2)
The given or available information may be the average (or mean) of the
variable, the standard deviation, the second or higher order moments, etc. The
entropy maximum principle states that the suitable probability density function
which describes the random variable distribution is obtained by maximizing the
entropy S. This leads to an optimization problem constrained by the known
information formulated as follow:
fi g i ( x) p X ( x)dx
i=1 to n,
where fi, represent the available information. For instance, if fi = mX (mean of the
random variable X), g(x) = x whereas when fi = m2 (second order moment), g(x)
= x.
This optimization problem can be solved by minimizing a Lagrangian
function H which includes (1+n) Lagrange multipliers (0, i, i =1 to n)
associated to the (1+n) constraints. The function H is given by:
,,
exp
(4)
,
an indicator
1ifx a, b
0ifnot
(5)
exp
(6)
3.2 Results
To illustrate this methodology, we focus on the rutting results obtained at 30 000
cycles. The average and standard deviation values of the rut for Hot and warm
bituminous mixtures are recapitulated in table 2.
Table 2:
Standard deviation_rut
HMA
4.23
1.62
WMA
6.53
0.74
In our case, given the average and the STD values and assuming that the
random variable associated to the rut depth shall not take nor negative neither
infinite value, the minimization of Eq. (x) leads to a Gaussian distribution:
pRut ( r ) exp( 0 1r 2r 2 )
(7)
Materials Characterisation V
(a)
Figure 4:
313
(b)
Figure 5:
4 Conclusion
A series of experimental tests were performed in this paper to compare the
rutting performance of HMA and WMA obtained by the incorporation of surface
WIT Transactions on Engineering Sciences, Vol 72, 2011 WIT Press
www.witpress.com, ISSN 1743-3533 (on-line)
References
[1] International Technology Scanning Program, Warm-Mix Asphalt:
European Practice, 2008.
[2] Brosseaud Y., Ecologiques, scuritaires, confortables, les enrobs de
demain se feront autrement : Prsentation des enrobes tides, 2006.
[3] Collectif. Un sujet chaud, les enrobs basses calories. Bitume info. N12,
pp. 12-15, 2006.
[4] Brosseaud Y., Warm asphalt-Overview in France. LCPC, France,
Prsentation to WMA scan team, 2007.
[5] Brosseaud Y. & Saint Jacques M, Warm Asphalt Mixes: Overview of This
New Technology in France, Second European Road. Transport Research
Arena Europe, Ljubljana, Slovnie, 21-24, 2008.
[6] Harder, G. et al (2008). Energy and environmental gains of warm and halfwarm asphalt mix: quantitative approach. In: Transportation Research
Board, Washington D.C., 2008.
[7] Warm mix asphalt technologies and research, Government engineering,
July-August, 2007.
[8] Perraton D., Di Benedetto H., Sauzat C. et al., Rutting of bituminous
mixtures: wheel tracking tests campaign analysis. Materials and structures,
Online First, 2010.
[9] Eddhahak Ouni A., Dony A. & Colin J., Assessment of a probabilistic
parametric rheological model to predict the rutting resistance of bitumen,
ASMDO 2010.
[10] Cooley Jr., Allen L., Prithvi S. Kandhal, & M. Shane Buchanan. Loaded
Wheel testers in the United States: State of the Practice. Transportation
Research Board E-Circular number E-C016, 2000.
[11] Shannon, C.E., A Mathematical Theory of Communication, Bell System
Technical Journal 27, pp. 379423 & 623656, 1948.
[12] Jaynes, E.T., Information theory and statistical mechanics, Physical
Review 106, pp. 620-630, 1957.
Materials Characterisation V
315
Abstract
In order to ascertain the presence of very fine material passing the No. 200 sieve,
cleanliness tests are performed in the laboratory. The most important of these
tests refer to the determination of the quantity of fines, of the sand equivalent
value, of the plasticity index, as well as the methylene blue value of the material
tested. Performing the sand equivalent test the percentage of very fine dust in
claylike form can be established. Materials with very low sand equivalent are
characterized as impure and the possibility of the existence of clay size grains
in them is very high. In such cases it is necessary to perform a test in order to
determine the plasticity of the soil material. The determination of the plasticity
index of natural soils specifies their suitability as subgrade layers in pavement
structures. Soils having a sand equivalent value lower than 10% will develop
excessive swell with a simultaneous decrease in their bearing capacity when a
pavement is built on them. The quantity of methylene blue absorbed by clay
components of a soil mass is proportional to the specific area of the clay
minerals. The methylene blue test supplements the sand equivalent and Atterberg
limits tests, since with these the existence of clay-size grains is determined, but
not the existence of active clay minerals as well. A number of soil samples have
been examined in the laboratory, involving the procedures of cleanliness tests.
The quantity of fine grained material was determined and their activity was
accessed. The magnitudes were correlated in order to check the suitability of
soils as a pavement foundation. Empirical relationships have been established
which connect the attributes characterizing the cleanliness of soil materials.
These relationships compare very well with most of the findings published
worldwide.
Keywords: swelling soils, methylene blue test, sand equivalent, cleanliness.
WIT Transactions on Engineering Sciences, Vol 72, 2011 WIT Press
www.witpress.com, ISSN 1743-3533 (on-line)
doi:10.2495/MC110281
1 Introduction
In most pavement structure techniques, suitability is one of the essential
characteristics of soil materials. The term "suitability" characterizes the absence
of undesirable elements. These elements are soluble salts, iron hydroxides,
organic matter, as well as clays. They can present sickliness because of their
fineness, their physicochemical activity or because of their effect in the binding
materials. In view of their frequency of appearance in alluvial deposits and rocks,
their hydrophilic nature, and of their affinity and plasticity properties, clays are
the most harmful.
Swelling soils often cause serious damages in pavement structures, while, on
the other hand, they cannot be used as embankment material in roads. The
swelling potential of soils is mainly affected by their clayey fraction (Kollaros
and Athanasopoulou [1]). In order to ascertain the existence of very thin material
passing the No. 200 sieve cleanliness tests are performed. The more important
tests are consisted in the determination of the quantity of fines, of the plasticity
index, the equivalent sand and the methylene blue value.
The quantity of material passing the 0,075 mm sieve (#No. 200), expressed as
a percentage of the total material, is determined by the sieve analysis test, either
through the dry or, more usually, through the wet process, according to
AASHTO T88 or ASTM D422 standards.
The plasticity index, which is defined as the difference in contained moisture
between the liquid and plasticity limits: PI=LL-PL, it is a value that helps to
recognize soil characteristics and to properly classify them. The plasticity index
determines the appropriateness of soils to be used as subgrades in roadway
foundations. Soils having a high PI value, such as clayey, silty and some sandsilt soils, are inadequate for the foundation of pavements. In Table 1 the degree
of plasticity is given in terms of PI and of soil characteristics in a dry condition,
for the corresponding scale of the PI values.
The plasticity index is useful for materials containing high percentages of silt
and clay, while in the region not covered by the PI use is made of the sand
Table 1:
PI
>35
16-35
7-15
4-6
0-3
Soil
Highly
Plastic
Plastic
Medium
plasticity
Slightly
plastic
Non Plastic
Materials Characterisation V
317
equivalent value. The sand equivalent, SE, test arose in view of the need for a
quick quality indication of paving materials. This test was devised by Hveem in
1953, it was put in use in 1956 and it has been adopted for laboratorial use since
1966.
More specifically, with the sand equivalent test the proportion of very fine
dust of claylike form is determined. For a SE<20 value, the plasticity index
should be determined. In the region of higher SE values, the plasticity index is
not determined (Non Plastic) and therefore it cannot replace the sand equivalent.
On the basis of comparisons of sand equivalent values to the results of other
tests it has been found that most of the soils presenting high swell potential when
saturated with water, is possible to be recognized by the SE value. Soils on top of
which a pavement is to be constructed develop excess swell, with a simultaneous
decrease of their bearing capacity in saturated conditions, provided that the SE
value is lower than 10%. The different fine grained materials could be classified
in three groups, as shown in Table 2, depending on their sand equivalent value.
Table 2:
Sand Equivalent
High Values
Medium Values
Low Values
Fine-grained materials
Clean non-cohesive sands
Sands blended with some quantities of silt and clay
Clays and silt-clayey mixtures
Soil Property
Clay content (material < 2) (%)
Percentage passing the No. 200 sieve (%)
Liquid Limit (%)
Plasticity Limit (%)
Plasticity Index (%)
Linear shrinkage (%)
Volume of dye, Vd (cm3)
Sand equivalent, SE
Skempton Activity [PI/(%material<2)]
Mean Value
41.42
74.65
52.73
22.24
29.73
12.93
20.35
14.46
0.71
Range of values
954
29.2390.13
2976
1729
1747
10.015.7
627.5
533
0.581.00
Materials Characterisation V
319
clays in the region fall in group A-7 according to the AASHTO classification
method. This group involves soil materials poor or unsuitable to be used as
foundation.
40
30
20
10
20
30
40
50
60
Figure 1:
(1)
PI=1,201Vd+5.79
(2)
In particular, a high degree of correlation has been obtained for the methylene
blue value and the percentage of claylike fraction in the samples tested. The
linear regression eqn. (3) connecting these attributes gave a coefficient of
determination, R2, equal to 0,82.
%material< 2=2.019Vd+0.34
(3)
(4)
50
40
30
20
10
10
12
14
16
18
20
22
24
26
28
Vd (cm )
Figure 2:
30
Materials Characterisation V
321
60
40
20
0
5
10
15
20
25
30
Vd (cm )
Figure 3:
100
80
60
40
20
5
10
15
20
25
30
Vd (cm )
Figure 4:
100
80
60
40
20
0
10
15
20
25
30
35
Sand Equivalent, SE
Figure 5:
In eqn. (5) the sand equivalent value, SE, is the independent variable and the
percentage of material passing the sieve No. 200 is the dependent variable. This
equation yielded a coefficient of determination, R2= 0,66.
%material< No. 200=1.994 SE+103.48
(5)
The sand equivalent values have been related to the respective values
obtained by the testing of the samples using the methylene blue test. The
equation of the curve that fits to the data in figure 6 has a coefficient of
determination, R2= 0,43 and is depicted in eqn. (6).
Vd=-0,587 SE+28.98
(6)
The negative gradient of the regression lines in figures 5 and 6 means that an
increase of the sand equivalent entails a reduction in the respective values of both
the fine-grained material and the quantity of dye absorbed by the clay particles.
Empirical relations similar with those resulted from the statistical processing
of the laboratorial results, have been reported in the international bibliography.
Many research efforts have been devoted to the correlation of the Atterberg
limits with the attributes tested and with the cation exchange capacity, CEC.
High level of correlation (r=0,80) has been obtained between the plasticity index
and the CEC (Taylor [3]). Also, a coefficient R2=0,56 was found for the relation
of the Skempton activity (PI/% of material<2) with the cation exchange
WIT Transactions on Engineering Sciences, Vol 72, 2011 WIT Press
www.witpress.com, ISSN 1743-3533 (on-line)
Materials Characterisation V
323
30
Vd (cm3)
26
22
18
14
10
0
10
15
20
25
30
Sand Equivalent, SE
Figure 6:
capacity (Sweere and Galjaard [4]). The plasticity index of 70 Illinois clay
samples (Odell et al. [5]) was fairly correlated with the content of material<2
through the eqn. (7):
PI=0,568 (%material<2)+1.09
(7)
A very good correlation between the apparent surface area calculated by the
methylene blue test and the percentage of particles having a diameter smaller
than 1 has been found by Xeidakis [6]. In this research 8 clayey minerals had
been tested and a correlation coefficient r=0,946 was found. A relatively weaker
correlation (r=0,65) has been reported for 19 British soils (Farrar and Coleman
[7]) which contained some montmorillonite.
Nikolaides et al. [8] published the results of SE and MB values obtained at 16
samples from Greek quarries. They found that there is no correlation between
MB and SE values.
The results of a series of tests performed on aggregates from production sites
and 45 mixtures prepared in the laboratory to clarify the effect of potentially
harmful fines on the MB and SE values are described by Petkovek et al. [9].
Particular attention is paid to the prediction of approximate limits for SE and MB
values, taking into account the specific geological conditions of the aggregate
sources and past experience using the same aggregates.
4 Conclusions
Swelling clay soils are abandoned in the region of Thrace, Northern Greece,
presenting many risks for the construction of pavements. In a new classification
system, it is possible for soils with homogeneous technical attributes to be
distinguished in categories, which they will be based on the activity of the clayey
phase -more specifically on the total specific area, that is, on the methylene blue
value- in combination with their grain-size gradation.
The methylene blue test has been found to be an easy and very repeatable test
method. Because of their simplicity, the suitability controls which are supported
by the laboratorial tests, can be also conveyed in field testing.
The correlation of values of attributes obtained from the tests in methylene
blue with the respective values of the plasticity index, the soil gradation (clay
fraction, percent of material passing the No. 200 sieve) and with the sand
equivalent showed that linearity exists in these relations and that the correlation
coefficients were particularly high. For the relations to merit general application,
they should be confirmed with the repetition of experimental processes in many
more samples from the region of Thrace.
References
[1] Kollaros, G. & Athanasopoulou, A., The character and identification of
swelling soils in road construction projects, Proc. International Symposium
on Engineering Geology and the Environment IAEG, eds. Marinos, Koukis,
Tsiambaos & Stournaras Balkema: Athens, pp. 187-192, 1997.
[2] Tran, N.L. & Millon-Devigne, P., L'essai au bleu de mthylne en
turbidimtrique, Bulletin de lAssociation Internationale de Gologie de
lIngnieur, 29(1), pp. 453-456, 1984.
[3] Taylor, R.K., Cation exchange in clays and mudrocks by methylene blue. J.
Chem. Tech. Viotechnol., 35A, pp. 195-207, 1985.
[4] Sweere, G.T.H. & Galjaard, P.J., The methylene blue test as a rapid means
of estimating the cation exchange capacity of soils. International Symposium
of Geotechnical Aspects of Mass and Material Transportation, Bangkok, pp.
47-52, 1984.
[5] Odell, R.T., Thornburn, T.H. & McKenzy, L., Relationships of Atterberg
limits to some other properties of Illinois Soils. Proceedings of the Soil
Science Society of America, 24(5), pp. 297-300, 1960.
[6] Xeidakis, G.S., Assessment of the engineering and other properties of
expansive soils by various methods. Ph.D. Thesis, Dept. of Civil
Engineering, University of Leeds, England, 407 p., 1979.
[7] Farrar, D.M. & Coleman, J.D., The correlation of surface area with other
properties of 19 British clay soils. Journal of Soil Science, 18(1), pp. 118124, 1967.
Materials Characterisation V
325
[8] Nikolaides, A., Manthos, E. & Sarafidou, M., Sand equivalent and methylene
blue value of aggregates for highway engineering. Foundations of civil and
environmental engineering 10, Publishing House of Poznan University of
Technology, Poznan, pp 111-121, 2007.
[9] Petkovek, A., Maek, M., Pavi, P. & Bohar, F., Fines characterization
through the methylene blue and sand equivalent test: comparison with other
experimental techniques and application of criteria to the aggregate quality
assessment, Bull Eng Geol Environ 69, pp. 561574, 2010.
Materials Characterisation V
327
Abstract
Highway construction engineers often face the need for more stable, durable and,
at the same time, more economic road structures. This is nowadays true because
of increased traffic volumes and heavier loads on the roadways. As a
consequence, enhanced pavement structures and improved subgrades is a
necessity. The international highway New Egnatia crosses areas in Thrace,
Northern Greece with abundant clayey soils having poor technical properties.
The treatment of physical soils with some substances could bring up new
materials, which would operate better under the traffic and environmental
conditions. This has led to the decision to investigate the possibilities of
improving the existing soil materials using chemical additives. Soil samples were
collected from the abovementioned area and mixed with lime and fly ash, in
various proportions. The modification of the soil properties with special
emphasis on their strength has been examined in the laboratory after different
curing periods.
The experimental results have shown that the unconfined compressive
strength increased as a function of both the percentage of additive in the mixture
and the time of curing. The improvement depended upon the soil mineralogy and
the kind and quantity of exchangeable cations. This holds true when the
influence of the kind and quantity of the additive is taken into account. The
effects of lime on the swelling clayey soils tested were more beneficial than
those of fly ash.
The soils after their treatment could be used as a subgrade or even as a
subbase layer in roadway pavements.
Keywords: stabilization, lime, fly ash, engineering properties, unconfined
compressive strength, optimum moisture.
WIT Transactions on Engineering Sciences, Vol 72, 2011 WIT Press
www.witpress.com, ISSN 1743-3533 (on-line)
doi:10.2495/MC110291
1 Introduction
The continuous growth of the traffic volume and of vehicle's size, has made
obvious the need for more stable, durable and, at the same time, more economic
road structures. Consequently, a requirement exists for improved properties of
the pavement structure as well as of the physical subgrade. The treatment of
physical soils and base materials with some substances could bring up new
materials, which would operate better under the traffic and environmental
conditions. In this procedure, called soil stabilization, such substances as
Portland cement, bituminous materials, lime, fly ash or alkali salts could be used.
It has been recognized by numerous investigators (Bell [1], Zhang and Cao
[2]) that the addition of lime and/or fly ash to soil materials may cause various
beneficial changes to the engineering properties of fine-grained soils, such as the
reduction of plasticity, the reduction of shrinkage-swelling potential and the
improvement of strength characteristics.
Roadways have a high potential for large volume use of the fly ash stabilized
soils. Arora and Aydilek [3] investigated the use of Class F fly ash amended soilcement or soil-lime as base layers in highways. Unconfined compression, CBR,
and resilient modulus tests were conducted. Required base thicknesses were
calculated using the strength parameters. The strength of a mixture is highly
dependent on the curing period, compactive energy, cement content, and water
content at compaction. Lime treatment didnt provide sufficient strength for
designing the mixtures as highway bases.
The strength of soil-lime-fly ash mixtures could be estimated by various
methods such as unconfined compression, CBR, the Hveem stabilometer and
triaxial tests. The most commonly used method is the unconfined compression
test, not necessarily being the most appropriate for all purposes. The strength of
soil-lime or soil-fly ash mixtures depends on many variables such as the soil
type, the lime and fly ash content, the additive type, the time and method of
curing (temperature and humidity), the water content, the unit weight and the
time interval between mixing and compaction (oka [4]).
The addition of lime, fly ash, and lime/fly ash to three clayey soils led to a
reduction of the plasticity index and contributed to an increase in the optimum
moisture content and a decrease in the maximum dry density (Hesham [5]). The
optimum lime content ranged from 3 to 5%, while the optimum fly ash content
between 16 and 35%. The optimum lime/fly ash content for the three soils was
(2.5%L+8%FA), (2%L+12%FA) and (3%L+20%FA). The UCS, Esecant, CBR,
and the Velocity of ultrasonic p-waves, Vp, values increased slightly with an
increment of the dry density of the untreated compacted soils (due to the
compaction process) and strongly due to the addition of the chemical stabilizing
agents (lime, fly ash, and lime/fly ash) whereas the formed cementitious
compounds (as a result of the chemical reactions between the silica and the
alumina and the additives) joined the soil particles.
Basic studies on pozzolanes and on the pozzolanic properties of fly ash have
been carried out by Adu-Gyamfi [6] and Brooks [7]. Considering the nature of
the lime used in the mixtures, many investigators such as Ingles and Metcalf [8],
WIT Transactions on Engineering Sciences, Vol 72, 2011 WIT Press
www.witpress.com, ISSN 1743-3533 (on-line)
Materials Characterisation V
329
have shown that calcarious limes gave higher strength results than those given by
dolomitic limes.
In Greece, many roads have been built using stabilization techniques for the
subbase materials. The behaviour of these roads has been judged as fairly
satisfactory. The scope in some of these trials was the use of large quantities of
fly ash per surface unit, in order to take advantage of the surplus of this material.
However, the cost of transporting the material from places of its production or
deposition to the areas of the projects under consideration, which are located to a
relatively small distance from the power stations, is a very restrictive factor.
The Research Centre of Public Works (KEDE) has carried out significant
studies on the stabilization of aggregates and clayey soils, mainly with Portland
cement and with lime and fly ash since 1982. The exploitation of Megalopolis
fly ash and its applications in highway construction in adjacent areas has been
examined by Marsellos et al. [9]. Further studies have been undertaken by Greek
Universities, the Technical Chamber of Greece, and other researchers.
The main objective of this work is to test the capability of lime and fly ash in
improving the engineering properties of clayey soils from the areas of Thrace, in
order to use them as stabilized layers in road construction and to apply more
economic processes in constructing new pavements or improving existing ones.
Loss on Ignition
SiO2
Al2O3
Fe2O3
CaO
MgO
K2O
Na2O
Properties
Specific gravity (kg/cm2)
Liquid Limit (%)
Plastic Limit (%)
Plasticity Index (%)
Linear Shrinkage (%)
Free Swell Index (%)
Maximum Dry Density (kg/m3)
Optimum Moisture Content (%)
Grain Size Distribution
Sand and Gravels (%)
Silt (%)
Clay (%)
Classification
AASHO
USCS
Unconfined Compressive Strength (kg/cm2)
Soils
Black Soil (S1) Brown Soil (S2)
2.5
2.7
76
51
29
23
47
28
13.3
10
95
51
1588
1707
21.7
17.8
24.3
22.7
53.0
16.2
35.8
48.0
A-7-6
CH
1.5
A-7-6
CL
2.7
The basic properties of the soils were determined from representative samples
and for the soil fraction passing the o 40 (425 ) sieve. The liquid limit of these
soils was found using the Casagrande method.
The grain size distribution of the soils has been determined by both the dry
method (AASHTO T-27) and hydrometer analysis. The physical properties of
the soils studied and their unconfined compressive strength (UCS) are presented
in Table 2.
Both soils are classified as Group A-7-6 according to the AASHO
classification system, while, according to the Unified Classification System, are
classified as CH and CL respectively. Soil samples passing the No 4 (4.75 mm)
sieve were used in order to find the dry density-moisture content relation with
the standard Proctor compaction test. The soils were thoroughly mixed with
different moisture contents (14% to 30%) and were cured in a moisture room for
24 hours before they were compacted, for uniformity purposes. The maximum
dry density and the optimum moisture content are shown in Table 2.
The lime used in this study was a typical commercial hydrated calcitic lime,
having a high CaO content (65.25%). It was supplied by the AIMOS Lime
Company, Drama, Greece which has a 200 ton daily production. The chemical
composition of this lime is shown in Table 3.
The term fly ash represents the fine-grained ash residue produced from
pulverized coal combustion and carried away by the hot gases comes out of the
WIT Transactions on Engineering Sciences, Vol 72, 2011 WIT Press
www.witpress.com, ISSN 1743-3533 (on-line)
Materials Characterisation V
331
chimney. This residue is usually collected with appropriate filter put along the
chimney. The fly ash used for the preparation of the laboratory specimens was
supplied by the Ptolemaida Power Station (6,000,000 ton/year). The chemical
composition of the fly ash is shown in Table 3.
Table 3:
Properties
Loss on Ignition
SiO2
Al2O3
Fe2O3
CaO
MgO
K2O
Na2O
Fly ash
13.90 (%)
29.95 (%)
10.85 (%)
4.57 (%)
20.00 (%)
1.90 (%)
0.95 (%)
0.32 (%)
Alteration of properties of black clay treated with lime and fly ash.
Materials
Soil
g
100
100
100
100
100
100
100
100
100
100
F.A Lime
G
g
0
0
4
0
8
0
12
0
0
4
0
7
0
10
3
1
6
2
5
1
LL
%
76
69
64
59
68
60
55
66
65
59
Atterberg
Limits
PL
%
29
32
35
39
39
41
43
34
48
42
PI
%
47
37
29
20
29
19
12
32
17
17
Compaction
Characteristics
M.D.D. O.M.C.
%
Kg/m3
1588
21.7
1526
25.6
1487
26.4
1422
31.2
1453
28.8
1449
29.4
1445
29.9
1476
28.2
1453
29.4
1468
28.8
7
Days
1.5
4.3
6.4
7.6
6.0
6.5
7.4
3.3
7.9
3.8
UCS
(kg/cm2)
28
90
Days Days
1.5
1.5
4.5
5.6
7.2
7.9
8.8
9.5
9.6
13.6
11.7 22.9
13.5 16.4
4.1
5.0
9.3
11.7
4.3
4.3
Materials Characterisation V
Table 5:
333
Alteration of properties of brown clay treated with lime and fly ash.
Materials
Soil F.A. Lime
g
G
g
100 0
0
100 4
0
100 8
0
100 12
0
100 0
4
100 0
7
100 0
10
100 3
1
100 6
2
100 5
1
LL
%
51
50
47
45
49
47
42
46
43
40
Atterberg
Limits
PL
%
23
33
37
40
NP
NP
NP
43
42
39
PI
%
28
17
10
5
---3
1
1
Compaction
UCS
Characteristics
(kg/cm2)
M.D.D. O.M.C.
7
28
90
kg/m3
%
Days Days Days
1707
17.8
2.7 2.7
2.7
1611
22.1
3.8 4.3
4.5
1577
24.5
6.6 7.9
9.2
1477
28.1
8.8 9.5 12.4
1575
24.6
3.5 5.7
8.5
1559
25.1
3.9 5.8
9.0
1525
26.5
3.9 6.0
8.4
1604
22.6
5.3 6.1
7.1
1558
24.9
8.4 11.7 12.6
1573
23.9
6.3 7.5
8.9
The change of OMC and MDD was gradual when fly ash was used, whereas
with the admixture of lime a rapid change existed with small percentages of
additive and remained almost constant thereafter. This could be attributed to the
reaction rate between the clayey soil and lime, as well as to its quick flocculation
due to quick exchange of soil cations with Ca++ from the lime and depression of
the double layer. On the other hand, the end change of these properties is greater
with fly ash than with lime due to the reaction of the soil with the constituents of
fly ash other than CaO, like SiO2, Al2O3 and MgO.
Considering the strength change of the soils, the UCS increased both with the
percentage of the additive and with the time of curing as it is demonstrated in
figures 1 to 4. In the case of lime addition, a dramatic increase occurred in the
strength of the soil (more than 10 times) with addition of only 4% lime. This
high rate of increase in soil strength was reduced with the increase of lime
content and at some 8-10% of lime the UCS remained more or less constant or
started to decrease (point of soil satisfaction). This percentage is recognized as
the lime modification optimum (LMO) of the soil. The same trends hold true for
the fly ash, though the rate of increase was lower and there was no point of soil
satisfaction. The strength increased in an almost constant rate when the
percentage of fly ash in the mixture was increased. As it is shown in figures 1 to
4 the rate of strength gain and the ultimate strength were different for both the
additive and the soil. Lime proved to be much more effective than the fly ash in
the case of the soil S1; the opposite has been observed for the soil S2.
The difference in the soil behaviour is certainly due to the differences in the
mineralogy of the soil and the kind of the exchangeable cations present. The
brown soil S2 is almost saturated by Ca++ cations, therefore the addition of lime
has little effect on the exchangeable cations of the soil. It is well-known that lime
has a more pronounced effect on swelling clay minerals (illite, caolinite) due to
the greater depression of the double layer. For the same reasons, fly ash yielded a
little better results in soil S1 than in the brown soil S2.
WIT Transactions on Engineering Sciences, Vol 72, 2011 WIT Press
www.witpress.com, ISSN 1743-3533 (on-line)
24
20
90 D-L
16
28 D-L
12
8
7 D-L
90 D-FA
28 D-FA
7 D-FA
4
Additive: Fly Ash
Additive: Lime
0
0
Figure 1:
6
8
10
Percent Additive (%)
12
14
16
14
90 D-FA
12
10
28 D-FA
90 D-L
7 D-FA
8
6
28 D-L
7 D-L
2
Additive: Fly Ash
Additive: Lime
0
0
Figure 2:
6
8
10
Percent Additive (%)
12
14
16
Materials Characterisation V
335
25
Soil Sample S1
Additive: Fly Ash
Additive: Lime
Additive: Lime+Fly Ash
7%
20
10%
15
4%
2:6
10
12%
8%
4%
1:5
1:3
0%
0
7
28
49
70
91
Figure 3:
2:6
12%
10
8%
7%
10%
4%
1:5
1:3
5
4%
0%
0
7
28
49
70
91
Figure 4:
Materials Characterisation V
337
Considering the effect of the curing period, the results followed the same
general pattern as with the percentage of additive. That is, for soil S1, the effect
of time was more significant in the case of lime than in the case of fly ash,
whereas in soil S2 the UCS value raised much more with the curing time when
fly ash was added than in the case where lime was used.
This difference could also be attributed to the abundance of swelling minerals
in soil S1. The more the swelling minerals, the more lime is precipitated in the
clay surface and the more cementitious materials (CaCO3) are formed. The
increase of strength in swelling minerals is due rather to cation exchange
(flocculation of the clay) and to the cementitious reaction than to the pozzolanic
one (Xeidakis [13], Baykal et al. [14]).
Soil mixtures having limefly ash ratios of 13 and 26 have given a little
higher strength values than those of the mixture with each additive alone. So, the
26 limefly ash ratio resulted to a strength two times and 1.5 times greater than
that with the addition of fly ash alone and lime alone, respectively (figures 3 and
4). The soil-lime-fly ash mixtures exhibited final strength values intermediate to
those found for the soil-lime and soil-fly ash mixtures.
5 Conclusions
The admixture of lime and fly ash to two expansive clays have led to a
significant decrease of the liquid limit probably due to the depression of the
diffuse double layer thickness associated with the clay particles, the aggregation
of the clay and the coating by Ca(OH)2.
A progressive reduction in maximum dry density and increase in optimum
moisture content has been observed with the addition of these materials. The
decrease of maximum dry density of clay soils, after their treatment with lime
and fly ash, is an indication of the increase of the strength of the soil and the
increase of its bearing capacity.
The strength of the mixtures tested was much higher than that of the natural
soils, in all cases. In general, the strength of soil-fly ash and soil-lime mixtures
increased with an increase in the additive content, for all curing periods. For both
soils and additives, an increase in curing period resulted to an increase in
strength.
The increase of the UCS for the soil S1 was greater with the addition of lime
(up to 20 times greater than the original), than with the addition of fly ash. The
best results obtained when 7% lime was added to the soils and a 90 days curing
period followed the compaction of the specimens. This is attributed to less Ca++
and a greater percentage of clay minerals in this soil. The strength increase in
soil S2 was greater with fly ash than with lime. This may be due to a higher
content in Ca++ and caolinite, as well as to a lower content in swelling minerals.
The results showed that the mineralogy of the soil plays a decisive role in the
stabilization process and greatly affects the ultimate strength of the mixture.
The ultimate strength of the soil after its improvement is adequate for the soil
to be used as subgrade or embankment material in main roads, or even as
subbase in some secondary roads.
WIT Transactions on Engineering Sciences, Vol 72, 2011 WIT Press
www.witpress.com, ISSN 1743-3533 (on-line)
References
[1] Bell, F.G., An Examination of the Use of Lime and Pulverized Fly Ash to
Stabilize Clay Materials. Bulletin of the Association of Engineering
Geologists, XXX(4), pp. 469-479, 1993.
[2] Zhang, J. & Cao, X., Stabilization of Expansive Soil by Lime and Fly Ash.
Journal of Wuhan University of Technology - Mater. Sci. Ed. 17(4), pp.
73-77, 2002.
[3] Arora, S. & Aydilek A.H., Class F Fly-Ash-Amended Soils as Highway
Base Materials. J. Mat. in Civ. Engrg., 17(6), pp. 640-649, 2005.
[4] oka, E., Use of Class C fly ashes for the stabilization of an expansive
soil. Journal of Geotechnical and Geoenvironmental Engineering, 127(7),
pp. 568-573, 2001.
[5] Hesham, A.H.I., Treatment and improvement of the geotechnical properties
of different soft fine-grained soils using chemical stabilization. PhD.
Thesis, Martin Luther Halle-Wittenberg University, Germany, p. 182, 2006.
[6] Adu-Gyamfi, G., A Generalized theory for fly ash modified soils. Ph.D.
Thesis, Department of Civil Engineering and the Russ College of
Engineering and Technology, p. 247, 2006.
[7] Brooks, R.M., Soil stabilization with fly ash and rice husk ash.
International Journal of Research and Reviews in Applied Sciences, 1(3),
pp. 209-217, 2009.
[8] Ingles, O.G. & Metcalf, J.B., Soil Stabilization, Butterworths, Melbourne,
1972.
[9] Marsellos, N., Christoulas, S. & Kolias, S., Use of Fly Ash In Road
Construction. KEDE Bulletin, 3-4, Athens, 1986.
[10] Sridharan, A., Rao, S.M. & Murthy, N.S., Liquid Limit of Montmorillonite
Soils. ASTM Geotechnical Testing Journal, 9(3), pp. 156-159, 1986.
[11] Akoto, B.K.A., Influence of Flyash on the Strength Characteristics of LimeLaterite Soil Mixtures. Australian Road Research 18(4), pp. 224-231, 1988.
[12] Athanasopoulou, A., Improvement of the Mechanical Properties of
Materials Used in Earthworks and Pavements. Application to Soils
Encountered in the Area of Thrace, Ph.D. Dissertation, Department of Civil
Engineering, Democritus University of Thrace, p.569, 1995.
[13] Xeidakis, G.S., Assessment of the engineering and other properties of
expansive soils by various methods. Ph.D. Thesis, Dept. of Civil
Engineering, University of Leeds, England, 407 p., 1979.
[14] Baykal, G., Arman, A. & Ferrell, R., Accelerated Curing of Fly Ash-Lime
Mixtures. Transportation Research Record, 1219, pp. 82-92, 1989.
Materials Characterisation V
339
Abstract
Creep is an important factor that contributes to the load loss and tightness failure
of bolted joints. Retightening of the joint can be expensive, time consuming and
therefore is an undesirable solution. Currently most efforts are focussed on
reducing load losses directly by tightening to yield, improving material creep
properties or making joints less rigid. An alternative solution of current interest
is the use of bolts in shape memory alloy (SMAs). However, very few
experimental studies are available that demonstrate its feasibility. The objective
of this study is to exploit the benefit of the shape memory and superelasticity
behaviors of a SMA stud to recover the load losses due to creep and thermal
exposure of a gasket in a bolted joint assembly. This paper explores several
avenues to investigate and model the thermo-mechanical properties of a bolted
joint with a Nickel-Titanium SMA stud. A stiffness-based analytical model
which incorporates the Likhachev model of SMA is used as a representation of
an experimental bolted joint assembly. Using this model the rigidity of the
experimental setup is optimized to make the best use of the SMA properties of
the stud. This theoretical model is validated by a Finite Element (FE) Model
using a custom FE material model which also implements the SMA material
model. Finally an experimental test bench with an optimized stiffness derived
from analytical simulations is used, with and without gaskets to demonstrate the
ability of the SMA stud to recover load losses.
Keywords: shape memory alloys, bolted joints, creep, superelasticity, SMA.
1 Introduction
Load losses due to creep in any bolted joint can be problematic, even small creep
losses of 0.1mm, can cause a total loss of bolt load. Several methods are in use to
WIT Transactions on Engineering Sciences, Vol 72, 2011 WIT Press
www.witpress.com, ISSN 1743-3533 (on-line)
doi:10.2495/MC110301
2 Analytical model
In order to make best use of the shape-memory and super-elasticity effects of the
nickel-titanium stud in the experimental test apparatus, a simplified analytical
model was implemented.
The model consists of 4 elements: The SMA stud, a spring element to
represent the stiffness of the test rig, a rigid element which is used to simulate
Materials Characterisation V
341
the thermal expansion of the flange, and a creep element to simulate the
properties of the gasket.
Super-Elasticity
SMA Rod
Force applied
Force applied
Bent rod at low
temperature
Releasing force
returns rod to
undeformed shape
Figure 1:
Creep Element
Figure 2:
Analytical model.
Figure 3:
As shown in Fig. 2, there is considerable loss in bolt load during the heating
phase, before the austenitic transformation temperature. This is primarily caused
by the properties of the shape memory alloy stud. These properties can vary
considerably based on the heat treatment given and the nickel titanium ratios [9].
Using the linear creep model, a gasket creep displacement of 1 mm was
introduced during heating in a joint having an SMA bolt and compared to a joint
with a B7 steel bolt. It is clear that for large creep such as those expressed with
PTFE gaskets in industrial flanges, the SMA stud retains much higher loads than
the steel bolt.
3 Numerical FE model
The axisymmetric arrangement of the simulated flange with an SMA stud in the
center lends itself well to a 1 dimensional analysis which can be treated
analytically without great difficulties. However in order to model more complex
flanges and configurations with multiple bolts, and various non symmetric
WIT Transactions on Engineering Sciences, Vol 72, 2011 WIT Press
www.witpress.com, ISSN 1743-3533 (on-line)
Materials Characterisation V
343
Figure 4:
4 Experimental setup
The experimental test rig allows a gasket to be placed between two hollow
cylinders with a B7 steel bolt or an SMA bolt in the center. A cartridge heater is
placed around the assembly and the exposed surfaces are insulated to maintain an
even temperature distribution. The top flange is instrumented with a full strain
gauge bridge, calibrated to measure axial load. The SMA bolt is instrumented
with thermocouples and strain gauges, to ensure even temperature distribution
and measure the strain. Thermocouples are also used to measure air temperature
to control the heater and the temperature of the bolt.
WIT Transactions on Engineering Sciences, Vol 72, 2011 WIT Press
www.witpress.com, ISSN 1743-3533 (on-line)
Stress(MPa)
ANSYSModel
MATLABModel
50
100
150
200
Temperature(C)
Figure 5:
Thermocouples
CenteringPlates
TopFlange
Thermocouple
StrainGaugesfor
LoadDetermination
GasketPostion
BottomFlange
CenteringPlate
BottomPlate
SMA/B7Rod
Figure 6:
Strain Gauges
Materials Characterisation V
400
5%elongation
3%elongation
345
7%elongation
Stress(MPA)
300
200
100
0
0
20
Figure 7:
40
60
80
100
120
Temperature(C)
140
160
180
200
30
NiTiSMABolt
B7SteelBolt
25
GasketStress(MPa)
20
15
10
5
0
0
Figure 8:
20
40
60
80
100
Temperature(C)
120
140
160
SMABolt
LikachevModel
Stress(MPA)
500
400
300
200
100
0
0
Figure 9:
50
100
Temperature(C)
150
200
6 Conclusion
In conclusion, preliminary results are promising and the use of a SMA rod has
considerably reduced the load loss due to creep in the gasket. It was also
demonstrated that additional load can be generated to further compensate creep
losses at higher temperatures. The application of the Likachev model of SMA in
both the analytical and FEM models yields good congruence with experimental
results.
References
[1] Peairs, D.M., Gyuhae Park; Inman, D.J., 2004, Practical issues of
activating self-repairing bolted joints, Smart Materials and Structures, v
13, n 6, p 1414-23,
[2] Antonios, C, Inman D.J, Smaili, A., 2006, Experimental and Theoretical
Behavior of Self-healing Bolted Joints, Journal of Intelligent Material
Systems and Structures, v 17, n 6, p 499-509,
[3] Hesse, T., Ghorashi, M., Inman, D.J., 2004, Shape memory alloy in
tension and compression and its application as clamping-force actuator in a
bolted joint, Part 1 experimentation, Journal of Intelligent Material
Systems and Structures, v 15, n 8, p 577-87
[4] Labrecque, C., Braunovic, M., Terriault, P., Trochu, F., Schetky, M., 1996,
Experimental and theoretical evaluation of the behavior of a shape
memory alloy Belleville washer under different operating conditions,
Electrical Contacts, Proceedings of the Annual Holm Conference on
Electrical Contacts, p 195-204.
Materials Characterisation V
347
[5] Ma, Hongwei, Wilkinson, T., Cho, C., 2007, Feasibility study on a selfcentering beam-to-column connection by using the superelastic behavior of
SMAs, Smart Materials and Structures, v 16, n 5, p 1555-1563
[6] Ma, Hongwei, Cho, C., 2007, Application of superelasticity of SMAs in
bolted end-plate connection, Key Engineering Materials, v 353-358, pt.4,
p 3039-42.
[7] V.A. Likhatchev, V.G. Malinin, Structure-Analytical Theory of Strength,
Nauka, St-Petersburg, 1993 (in Russian).
[8] Therriault, P., Viens, F., Brailovski, V., 2006, Non-isothermal finite
element modeling of a shape memory alloy actuator using ANSYS,
Computational Materials Science, v 36, n 4, p 397-410.
[9] Brailovski, Vladimir, Prokoshkin, Sergei D., Khmelevskaya, Irina Yu.,
Inaekyan, Karine E., Demers, Vincent, Dobatkin, Sergei V., Tatyanin,
Evgeny V., 2006, Structure and properties of the Ti-50.0 at%Ni alloy after
strain hardening and nanocrystallizing thermomechanical processing,
Materials Transactions, v 47, n 3, p 795-804
[10] ANSYS, 2003, ANSYS Standard Manual, Version 11.0.
[11] Matlab, 2007, Version 7.4.0.287 (R2007a).
Section 7
Thermal analysis
Materials Characterisation V
351
Abstract
Adhesively bonded scarf repairs are the preferred method of repairing modern
composite structures as they provide high strength restoration and aerodynamic
flushness. Curing of the adhesive bondline is carried out by locally heating the
repair area. To assess repair design and heating practices simulation can be used
to model both the transient heat transfer during curing and the level of cure likely
to be achieved at different regions in the adhesive joint. In this paper a 3D curing
model is described to simulate heat transfer through a composite component. The
cure kinetics of a commercial epoxy resin adhesive have been determined using
isothermal Differential Scanning Calorimetric (DSC) analysis. Using these
kinetics the model is able to determine the influence of the exothermic reaction
within the adhesive on the overall temperature variation within the component.
An experimental programme has been carried out where composite material and
bonded repair patches have been cured with thermocouples providing measured
temperature/time data during the cycle. The results from the cure model are then
validated by comparison with these experimental results. The cure model is
capable of being used to optimise the cure cycle for a bonded repair, ensuring the
maximum degree of cure of the adhesive with minimum variation of temperature
within the bond line.
Keywords: cure kinetics, DSC analysis, numerical model.
1 Introduction
The increase in use of composites in the aerospace, automotive and civil
engineering industries has led to the increasing demand for development of
WIT Transactions on Engineering Sciences, Vol 72, 2011 WIT Press
www.witpress.com, ISSN 1743-3533 (on-line)
doi:10.2495/MC110311
2 Model description
A flat plate with dimensions of 560mm x 560mm with a uniform thickness of
10mm with a simple 120C cure cycle was simulated. The repair area consisted
of a patch of 260mm diameter with 5mm thickness and a 3 scarf angle.
The patch is bonded into the repair area using the same commercial film
adhesive as used in the DSC experiments. The cure cycle for this film adhesive is
provided by the manufacturer.
The equations were solved using a Gauss-Seidel algorithm where the
temperature and degree of cure were determined at each time step.
The cure kinetics equations in this work were based on the Cheung et al. [10]
paper and the heat transfer methodology is based on the work of Patankar [11]
entitled Numerical heat transfer and fluid flow.
WIT Transactions on Engineering Sciences, Vol 72, 2011 WIT Press
www.witpress.com, ISSN 1743-3533 (on-line)
Materials Characterisation V
353
(3-9)
The values of aE to aB represent the heat conductance between the grid point P
and the corresponding neighbour in the model where x, y and z represent the
dimensions of the control volume. The variable b is a constant consisting of the
internal energy and the rate of heat generation resulting from the term SC. This
term is given as:
where Q is the heat flow per unit mass (J/kg)
(10)
The harmonic means of the thermal conductivities were used to account for
the different materials used.
Finally the value of
TP is defined:
(11)
(12)
This rate of reaction is given by the well established Kamal and Sourour [12]
equation:
1
WIT Transactions on Engineering Sciences, Vol 72, 2011 WIT Press
www.witpress.com, ISSN 1743-3533 (on-line)
(13)
Figure 1:
Materials Characterisation V
355
The equation for the rate of cure has been given as:
(15)
where K1 and K2 are rate constants and m and n are reaction orders.
Further work using the DSC measurements showed the maximum degree of
cure does not reach 100% as towards the latter stages of the reaction it becomes
diffusion controlled. The current model was based on a final 100% cure for the
adhesive. Some models [14] have used a temperature dependent function max as
the maximum degree of cure to account for the diffusion controlled process in
the final stages of the curing reaction. This modification of the Kamal and
Sourour [12] equation is as follows:
(16)
This forms the basis for model 2 cure kinetics (see Table 1).
The least square best fit method can be used to minimise this difference
between the calculated and measured values by altering the values of K1, K2, m
and n. This produces fitted values of K1, K2, m and n for all temperatures which
can be used to find the cure kinetics of each model.
Figure 2:
Arrhenius found that by taking the natural logarithm (ln) of both K1 and K2
and plotting them in a graph against the inverse of temperature (1/T) two linear
correlations are produced, with gradient and intercept of (-E/R) and A
respectively [15]:
ln
ln
(17)
A2 (sec-1)
5
8.2985x10
3.435x1029
1.8294x10
12
4.8042x106
Cure kinetics.
E1 (J/mol)
E2 (J/mol)
72203
105960
1.1655
3.0108
257126
64296
0.8892
1.5493
These values were placed in the model to determine the effect of the
exothermic reaction during cure on the overall heat transfer through a composite
component.
Figure 3:
Materials Characterisation V
Table 2:
Model
No.
1
2
3
4
Cure
Temp
(C)
120
120
120
120
Thickness
(mm)
10
20
10
10
357
Model variations.
Ambient
Temperature
(C)
25
25
50
25
Rate
(C/min)
HTC
(Wm-2K-1)
1
1
1
3
15
15
15
15
lost through the edges of the panel. However, the effect of ramp rate on the
curing of the adhesive was significant. The increased ramp rate produced a far
higher degree of cure of the adhesive during the early stages of the cure cycle.
Figure 4:
Figure 5:
5 Experimentation
The model is based on the conduction of heat from a source above the
component. This is representative of a modern heating method used specifically
for repair of large composite components or for in-service applications. This
method involves a hot bonder used as the heat source of a number of CFRP
composite components of different thickness and geometry. These results were
used to validate the model and to suggest areas that required further
investigation.
Thermocouples were embedded into a 900mmx900mmx10mm panel in the
bottom right hand corner of the panel to match the model. Thermocouples 1-16
were placed after plies 8, 16, 24 and 32 in the arrangement shown in figure 6.
Figure 6:
Materials Characterisation V
359
Figure 7:
For example, the thermal conductivity values for each material were
considered to be the same in all directions. In practice, the thermal conductivity
of carbon fibre in the transverse direction is approximately 25% of that in the
longitudinal and through-thickness directions [16]. However, the experimental
trials compare favourably with the results from the models.
WIT Transactions on Engineering Sciences, Vol 72, 2011 WIT Press
www.witpress.com, ISSN 1743-3533 (on-line)
7 Conclusions
A finite difference model was produced to simulate heat transfer through a flat
CFRP composite panel during the cure cycle of a bonded repair process. The
effects of heating rates, insulation and environmental temperature were simulated
and the results validated through experimental investigation.
Further modification of the model from experimental results has improved
both the accuracy of the model and the ability to uniformly heat a composite
component. The results from the model have provided insight into the nature of
heat transfer under different conditions.
The large variance in temperature between the centre and edge thermocouples
was demonstrated in both the model and experimentation.
The model provided more information as to where the heat loss is most
significant, and with further trials in combination with further modelling this
heat loss could be significantly reduced.
The exothermic reaction during the cure of the adhesive was determined to
have an insignificant impact on the overall heat transfer through the component.
The incorporation of a Titanium bolt increased the local temperature and
therefore degree of cure of the adhesive in the early stages of the cure cycle.
However, this became insignificant towards the latter stages of the cure.
Ongoing work will involve the optimisation of the cure cycle for bonded
repair, residual stresses and deformation, and void formation.
Acknowledgements
This work was carried out as part of the CONTOUR project in collaboration with
Airbus Operations Limited under Welsh Assembly Government contract
reference number HE 09 COL 1030.
References
[1] Tomblin, J.S., Salah, L., Welch, J.M., Borgman, M.D., Bonded Repair of
Aircraft Composite Sandwich Structures, Final Report, DOT/FAA/AR03/74, Office of Aviation Research, Washington D.C. 20591, February
2004.
[2] Baker, G., Bonded Composite Repair of Fatigue-Cracked Primary Aircraft
Structure. Composite Structures, 47(1-4), pp. 431-443, 1999.
[3] Charalambides, M.N., Hardouin, R., Kinloch, A.J. Matthews,
F.L., Adhesively-Bonded Repairs to Fibre-Composite Materials
I: Experimental. Composites Part A: Applied Science and Manufacturing,
29(11), pp. 1371-81, 1998.
[4] Yi, S., Hilton, H. H., Ahmad M.F., A Finite Element Approach for Cure
Simulation of Thermosetting Matrix Composites. Computers & Structures,
64(1-4), pp. 383-388, 1997.
Materials Characterisation V
361
[5] Stewart, I., Chambers, A., Gordon, T., The Cohesive Mechanical Properties
of a Toughened Epoxy Adhesive as a Function of Cure Level. International
Journal of Adhesion & Adhesives, 27(4), pp. 277-287, 2007.
[6] Yu, H., Mhaisalkar, S.G., Wong, E.H., Teh, L.K. & Wong,
C.C., Investigation of Cure Kinetics and Its Effect on Adhesion Strength of
Nonconductive Adhesives Used in Flip Chip Assembly. IEEE Transactions
on Components and Packaging Technologies, 29(1), pp. 71-79, 2006.
[7] Loos, A.C., Springer, G.S., Curing of Epoxy Matrix Composites. Journal of
Composite Materials, 17(2), pp. 135-169, 1983.
[8] Bogetti T.A., Gillespie, J.W., Two-Dimensional Cure Simulation of Thick
Thermosetting Composites. Journal of Composite Materials, 25(3), pp.
239-273, 1991.
[9] Zhu, Q., Geubelle, P.H., Tucker, C.L., Dimensional Accuracy of Thermoset
Composites: Simulation of Process-Induced Residual Stresses. Journal of
Composite Materials, 35(24), pp. 2171-2205, 2001.
[10] Cheung, A., Yu, Y., Pochiraju, K., Three-Dimensional Finite Element
Simulation of Curing of Polymer Composites. Finite Elements in Analysis
and Design, 40(8), pp. 895-912, 2004.
[11] Patankar, S.V., Numerical Heat Transfer and Fluid Flow, Taylor &
Francis: Oxfordshire, 1980.
[12] Kamal, M.R., Sourour, S., Kinetics and Thermal Characterization of
Thermoset Cure. Polymer Engineering & Science, 13(1), pp. 59-64, 1973.
[13] Standard Test Method for Arrhenius Kinetic Constants for Thermally
Unstable Materials Using Differential Scanning Calorimetry and the
Flynn/Wall/Ozawa Method, ASTM E 698-05, 2005.
[14] Lee, C.L., Wei, K.H., Curing Kinetics and Viscosity Change of a Two-Part
Epoxy Resin During Mold Filling in Resin-Transfer Molding Process.
Journal of Applied Polymer Science, 77(10), pp. 2139-2148, 2000.
[15] Lee, J.Y., Choi, H.K., Shim, M.J., Kim, S.W., Kinetic Studies of an Epoxy
Cure Reaction by Isothermal DSC Analysis. Thermochimica Acta, 343
(1-2), pp. 111-117, 2000.
[16] Mutnuri, B., Liang, R., GangaRao, H., Thermal Conductivity
Characterization of FRP Composites: Experimental. ANTEC, May 7-11,
Charlotte, NC, 2006.
Materials Characterisation V
363
Abstract
Hygrothermal performances of three types of building envelopes based on
autoclaved aerated concrete (AAC) provided with different thermal insulating
materials (expanded polystyrene, hydrophilic mineral wool, AAC with extended
thermal insulation capability) are compared. The simulations are accomplished
using the computer code HEMOT based on the finite element method. Results of
the simulation are the moisture and temperature fields across the building
envelope, which in combination with mechanical parameters present a sufficient
data source for service life analysis.
Keywords: computational analysis, coupled heat and moisture transport,
autoclaved aerated concrete, building envelopes, climatic conditions.
1 Introduction
Autoclaved aerated concrete (AAC) is a structural material which is commonly
used around Europe, particularly as it combines ease of construction with
excellent combination of its mechanical and thermal properties. However,
despite the very good thermal properties of AAC, it can be anticipated that with
the increasing demand for energy savings it will become a necessity to provide it
with thermal insulating system or at least very good thermal insulating plaster to
meet stringent conditions given by future thermal standards. The choice of
proper insulating material for AAC is though quite difficult with respect to
service life of the whole building envelope. As it was demonstrated in [13],
most common insulation materials cause extreme hygrothermal straining of
WIT Transactions on Engineering Sciences, Vol 72, 2011 WIT Press
www.witpress.com, ISSN 1743-3533 (on-line)
doi:10.2495/MC110321
2 Computational analysis
The computer code HEMOT [4] is based on the general finite element package
SIFEL [5]. As basic input parameters of the mathematical model, hygric, thermal
and basic physical parameters of used materials, construction detail, initial and
boundary conditions and time specification of simulation are required.
Description of all input parameters in more detail is given later.
In the computer simulations we focused on a comparison of hygrothermal
behavior of several building envelopes based on AAC provided with different
thermal insulating materials.
2.1 Mathematical model
Knzels mathematical model of heat and moisture transport [6] was used in the
simulations which can be formulated as
d v
div D grad p grad ps
d t
(1)
dH T
divgradT Lv div p grad p s
dT t
(2)
D Dw
d v
d
(3)
Materials Characterisation V
365
start-up building envelope we chose AAC without any external finish which
allowed us to get real image about hygrothermal performance of AAC layer itself
[7]. In the next simulation we provided AAC with hydrophilic mineral wool
(marked as Var. 1), expanded polystyrene (Var. 2) and Multipor Ytong produced
by Xella CZ (Var. 3). All these envelopes were provided from interior and
exterior side with Baumit MVR Uni plaster which is recommended for AAC
structures as external finish. On the material interface between mineral wool and
AAC an adhesive mortar layer was placed. Description of used materials in more
detail is given in next subsection. Scheme of construction detail including the
dimensions of each layer is shown in Figure 1. In all investigated variations we
focused on the hygrothermal conditions at points within the AAC layer and
external plaster just 2 mm under its external surface which can be considered as
characteristic position from the point of view of possible frost damage.
Figure 1:
AAC
P4-500
Mamu
t M2
mortar
[kg m-3]
%]
c [J kg-1 K1
]
[-]
500
dry [W m-1
K-1]
sat [W m-1
K-1]
[m2 s-1]
Expanded
polystyrene
Multip
or
1430
Hydrophi
lic
mineral
wool
71
50
125
Baumit
MVR
Uni
plaster
1402
80.2
1020
1510
3.0 9.7
42.6
1020
96.0
810
97.0
1300
12.4
4.3
50
0.114
0.481
0.043
0.040
94.2
2230
3500
1.9
10.9
0.047
44.4
1020 1780
4.5
12.4
0.443
0.454
2.022
0.246
0.560
0.166
1.380
Fig. 2
1.07e9
0.201
8.4e-6
2.10e-11
Fig. 2
1.59e-9
0.000046
0.001
0.0078
0.042
0.01846
whyg
[m3 m-3]
Material characteristics.
1.00E-05
1.00E-06
1.00E-07
1.00E-08
AAC P4-500
Multipor
1.00E-09
1.00E-10
0
0.05
0.1
0.15
0.2
0.25
0.3
3
0.35
0.4
Figure 2:
Moisture diffusivity.
Materials Characterisation V
367
temperature 21C (see Fig. 3) was chosen. The simulation started on 15th July
and took 6 years. The final results show data obtained for the last year.
Figure 3:
Boundary conditions.
3 Computational results
The results of computational simulations are summarized in a set of figures
which describe hygric and thermal performance of studied material of building
envelope during a reference year. In all figures we focused on moments when
moisture content and temperature reached certain limits simultaneously. In case
of moisture content this limit was the value of hygroscopic moisture content (see
Tab. 1), in case of temperature the freezing point of water. When these two
conditions are fulfilled, contained liquid moisture is getting frozen. This leads to
consequent damage of material.
The materials capability to resist to freezing of contained water is
characterized by its freeze-thaw resistance which can be measured under
laboratory conditions [12]. This was accomplished for AAC and its freeze-thaw
resistance was set to 25 cycles [13]. Durability of AAC can be then calculated as
quotient of freeze-thaw resistance and number of freezing cycles appearing in the
material in building envelope during a year.
3.1 AAC wall provided with hydrophilic mineral wool
Hygrothermal performance of AAC and external plaster of building envelope
provided with hydrophilic mineral wool is captured in Figures 4 and 5.
The temperature in AAC block 2 mm under its external surface is lower than
in other investigated envelopes, but freezing point of water is not reached and
level of moisture content is deeply in underhygroscopic range. That means that
AAC block in not threatened by effects of freezing water.
Hygrothermal performance of exterior plaster is shown in Figure 5. Although
the temperature drops below zero many times during a reference year, this does
not happen simultaneously with moisture increase. So, there is not any water
which can freeze and damage the structure of plaster. During a reference year,
the overhygroscopic moisture content is reached only five times and only in
summer months. So we can summarize, that even in the plaster there are not any
freezing cycles.
WIT Transactions on Engineering Sciences, Vol 72, 2011 WIT Press
www.witpress.com, ISSN 1743-3533 (on-line)
0.08
Temperature [K]
0.07
300
0.09
Moisture content [m /m ]
320
280
0.06
0.05
260
0.04
240
0.03
0.02
220
0.01
200
731
0.00
781
831
881
931
981
1031
1081
Time [days]
Figure 4:
0.08
Temperature [K]
0.07
300
0.09
Moisture content [m /m ]
320
280
0.06
0.05
260
0.04
240
0.03
0.02
220
0.01
200
1700
0.00
1750
1800
1850
1900
1950
2000
2050
Time [days]
Figure 5:
Materials Characterisation V
369
0.10
Temperature 273.15 K
Temperature
Hygroscopic moisture content
Moisture content
0.08
Temperature [K]
0.07
300
0.09
Moisture content [m /m ]
320
280
0.06
0.05
260
0.04
240
0.03
0.02
220
0.01
200
1700
0.00
1750
1800
1850
1900
1950
2000
2050
Time [days]
Figure 6:
Temperature 273.15 K
Temperature
Hygroscopic moisture content
Moisture content
320
300
0.09
0.08
0.06
0.05
260
0.04
240
Temperature [K]
Moisture content [m /m ]
0.07
280
0.03
0.02
220
0.01
200
1700
0.00
1750
1800
1850
1900
1950
2000
2050
Time [days]
Figure 7:
0.08
Temperature [K]
0.07
300
0.09
Moisture content [m /m ]
320
280
0.06
0.05
260
0.04
240
0.03
0.02
220
0.01
200
1700
0.00
1750
1800
1850
1900
1950
2000
2050
Time [days]
Figure 8:
0.08
Temperature [K]
0.07
280
0.06
0.05
260
0.04
240
0.03
0.02
220
0.01
200
1700
0.00
1750
1800
1850
1900
1950
2000
2050
Time [days]
Figure 9:
300
0.09
Moisture content [m /m ]
320
Materials Characterisation V
371
4 Discussion
The protection of AAC walls against effects of weather conditions and freezing
water in particular is certainly a necessity. It was proved already in the previous
work [7]. There are two ways how to achieve that. The AAC wall can be
protected against increase of moisture content which can be accomplished by
suitable waterproof modification of external surface or we can protect the wall
against low temperatures using one of many thermal insulating materials. The
second possibility seems to be more advantageous with respect to stringent
conditions given by thermal standards which have to be met.
The advantage of thermal insulation applied on AAC is the fact that together
with frost the AAC is protected against moisture penetrating from exterior as
well. However, the application of thermal insulation goes along with negative
effects. One of the most significant is extreme straining of external finishes,
which are applied on. This was already indicated in [13] and also results
obtained in this paper confirmed it.
The main reason can be seen in the low moisture diffusivity of used thermal
insulating materials (expanded polystyrene, Multipor). This leads to slowing
down the moisture transport from exterior plaster towards the interior and to
moisture increase in plaster subsequently. Considering the direct exposition of
plaster to weather conditions, temperature in particular, water saturated plaster
can be then easily damaged by effects of freezing of contained liquid moisture. It
can be spoken in general, the lower moisture diffusivity of thermal insulating
material is, the more freezing cycles in exterior plaster will appear. In our case,
there were 15 freezing cycles in plaster applied on expanded polystyrene and 10
freezing cycles in plaster applied on Multipor and that corresponds to values of
moisture diffusivity.
Opposite situation occurred when hydrophilic mineral wool has been used.
Thanks to high moisture diffusivity of hydrophilic mineral wool, all the received
moisture can be quickly transported away, so the plaster stays relatively dry and
does not contain the liquid moisture. On the other hand, moisture content of
AAC is higher, which is obvious, when Figures 4, 6 and 8 are compared.
Although the moisture content is still very low, the total amount of moisture
contained in whole building envelope provided with hydrophilic mineral wool is
higher than in other studied cases so that the thermal insulating capability of that
building envelope may possibly decrease.
5 Conclusion
According to the results obtained in this paper, protection of AAC block in
building envelopes against weather conditions is a necessity. It can be
accomplished in many ways; one of the best seems to be using thermal insulating
materials from exterior side. In this paper, computational simulation of building
envelopes provided with different types of thermal insulating materials has been
accomplished. The best results were obtained when hydrophilic mineral wool
was under consideration. The other variants with expanded polystyrene or
WIT Transactions on Engineering Sciences, Vol 72, 2011 WIT Press
www.witpress.com, ISSN 1743-3533 (on-line)
Acknowledgement
This research has been supported by the Czech Science Foundation, under grant
No. 103/09/0016.
References
[1] Madra, J., Ko, V., Vejmelkov, E., ern, R., Rovnankov, P. et al,
Influence of material characteristics of concrete and thermal insulation on
the service life of exterior renders. In: Fourteenth International Conference
on Computational Methods and Experimental Measurements, Algarve,
Portugal. Wessex: WIT PRESS, p. 13-23. ISBN 978-1-84564-187-0. 2009.
[2] Madra, J., Ko, V., Koreck, T., ern, R. et al. Computational analysis
of hygrothermal performance of building envelope under different climatic
conditions In: Thermophysics 2010. Brno: University of Technology,
p. 180-187. ISBN 978-80-214-4166-8. 2010.
[3] Jerman, M., Ko, V., Madra, J., Vborn, J., ern, R, Water and heat
transport parameters of materials involved in AAC-based building
envelopes In: 1st Central European Symposium on Building Physics. Lodz:
Technical University of Lodz, p. 39-45. ISBN 978-83-7283-367-9. 2010.
[4] ern R., Complex System of Methods for Directed Design and
Assessment of Functional Properties of Building Materials: Assessment and
Synthesis of Analytical Data and Construction of the System. CTN CTU in
Prague, 192 201. 2010.
[5] Kruis, J., Koudelka, T., Krej, T., Efficient computer implementation of
coupled hydro-thermo-mechanical analysis. In: Mathematics and
Computers in Simulations, vol. 80, no. 8, pp. 1578-1588. ISSN 0378-4754.
2010.
[6] Knzel, H.M., Simultaneous Heat and Moisture Transport in Building
Components, Ph.D. Thesis, IRB Verlag: Stuttgart, pp. 1135, 1995.
WIT Transactions on Engineering Sciences, Vol 72, 2011 WIT Press
www.witpress.com, ISSN 1743-3533 (on-line)
Materials Characterisation V
373
[7] Madra, J., Ko, J., Ko, V., Vborn, J., ern, R., Computational
prediction of hygrothermal conditions in innovated AAC-based building
envelopes. In: Advanced Computational Methods and Experiments in Heat
Transfer XI. Southampton: WIT Press, pp. 291-301. ISBN 978-1-84564462-8. 2010.
[8] Jiikov, M. and ern, R., Effect of Hydrophilic Admixtures on Moisture
and Heat Transport and Storage Parameters of Mineral Wool, Construction
and Building Materials, 20, 425-434, 2006.
[9] Vborn, J. Stanoven tepelnch parametr a smrtn pi vysychn
vybranch probetonovch tvrnic firmy H+H esk republika s.r.o. a
Xella CZ. In: Proceedings of the International Conference 15th
Construmat 2009. Praha: esk vysok uen technick v Praze, pp.
449-461. 2009.
[10] Fukov, L., Moisture Properties of AAC Blocks in dependence on the
Environment Focused on Problems with Durability. Diploma thesis. Praha:
Czech Technical University in Prague, 2009.
[11] Jerman, M., Vborn, J., ern, R., Tepeln a vlhkostn charakteristiky
novch probetonovch vrobk, In: Stavebn obzor, ro. 20, . 1, s. 7-11.
ISSN 1210-4027. 2010.
[12] SN EN 15304. Determination of the freeze thaw resistance of
autoclaved aerated concrete. Praha: Czech Standards Institute, 2007.
[13] Vborn, J., Testing of the Drying Shrinkage and Freeze - thaw Resistance
of Autoclaved Aerated Concrete. In: XII International scientific conference
Technick zazen staveb a energie budov. Brno: Akademick
nakladatelstv CERM, pp. 215-218. 2009.
Section 8
Recycled materials
Materials Characterisation V
377
Abstract
This paper presents the application of stereology methods to the description of
morphological properties of polyurethane nanocomposites for medical
applications. The study of the cross-section surface structure of the obtained
materials was performed by Atomic Force Microscopy. The volume of hard
phase agglomerate was used to evaluate the degree of phase separation of the
examined nanocomposites. The relationships between the domain agglomerate
characteristics and the properties of nanocomposites obtained from them were
analysed.
The results showed that nanocomposites with non-modified nanosilica
dioxide (SiO2) and nanosilica dioxide modified with NH2 groups differs from
polyurethane within the following properties: size and volume of the
agglomerates of the hard domains, biocompatibility, thermo-mechanical and
abrasive wear resistance.
Keywords: nanocomposites, polyurethane, structure, image analysis, biomedical
application
1 Introduction
Polyurethanes (PURs) and their nanocomposites are a versatile plastic material,
formulated to provide good biocompatibility, flexural endurance, high strength,
high abrasion resistance and processing versatility over a wide range of
applications [1]. The most common use in medical devices is in short-term
implants [2]. Polyurethane elastomers are linear segmented copolymers
WIT Transactions on Engineering Sciences, Vol 72, 2011 WIT Press
www.witpress.com, ISSN 1743-3533 (on-line)
doi:10.2495/MC110331
Figure 1:
Materials Characterisation V
379
2 Experimental
2.1 Materials
The following reactants were used in the synthesis of PURs:
polyoxythetramethylene glycol (PTMG) of molecular weight 2023 g/mol
(Therathane 2000) was supplied by Du Pont Nemours Co. ; 4.4diphenylmethane diisocyanate (MDI) purchased from Aldrich Chemical Co.
(Germany); ethylene glycol (GE) and glycerine (G) POCH (Gliwice, Poland).
Polyol was dehydrated by mixing under vacuum for 2h at 120C. Ethylene
glycol and glycerin was dried under a molecular sieve.
Nanosilica powder and nanosilica powder modified with NH2 groups, with
primary particle size of 75 nm were used as nanofiller (ICHP Warsaw). To
prepare the composites, nanosilica polyetherodiols 20% wt. concentrate was
made. Firstly, polyetherodiol was melted down in the oven under 80C degree.
Then nanosilica was put into it and mixed with ultrasonic homogenizer VCX
750 by Sonics during 30min in pulse mode 3/3 (3sec mixing, 3sec stop).
2.2 Polyurethane and nanocomposites synthesis
Segmented polyether based polyurethanes with substrates molar ratio PTMG:
MDI: GE: G equal to 40:80:27:24 (1:2:0.679:0.151), constant isocyanates index
of 1.05 and with hard segment contents 20 wt.% was synthesized using a onestep polymerisation method. The polyol was cooled to 70C3C, with glycol
added and glycerin blended for 5min. Then the mixture was cooled to 60C3C
and MDI added. The samples were obtained with free casting method. The heat
up process was carried out in the oven for 16 hours. To prepare the composites,
appropriate quantities of concentrate were added to the polyol. The mixture was
dehydrated in temperature of 120C5C under 2-5 hPa pressure. Next the
process was run in the same way as polyurethane synthesis. Description of
achieved materials is presented in Table 1.
Table 1: Description of the achieved materials.
Sample
Amount of nanofiller [wt.%]
Modification of nanofiller
0
0
-
1
0.5
-
2
1.0
-
3
2.0
-
1A
0.5
-NH2
2A
1.0
-NH2
3A
2.0
-NH2
2.3 Characterization
2.3.1 Dynamic mechanical analysis
Dynamic tests, by dynamic mechanical analysis (DMA) on a Thermal
Instruments dynamic mechanical analyzer (Q800 TA), were also carried out in
three-point bending mode on specimens with dimensions of 12 x 2 x 60mm3.
Tests were performed with the amplitude of deformation during the bending of
25 m. The frequency-dependent storage modulus was also evaluated with a
WIT Transactions on Engineering Sciences, Vol 72, 2011 WIT Press
www.witpress.com, ISSN 1743-3533 (on-line)
Vs
, 100%
Vt
where Vs is the loss of volume of the standard rubber (mm3) and Vt is the loss of
volume of the test sample (mm3).
The density figures for the test pieces, which were necessary for calculations,
were found by the method described in the standard PN-ISO [14].
2.3.3 Atomic Force Microscopy (AFM)
AFM images were recorded at 37C in air using a Digital Instrument Multimode
Nanoscope V (Digital Instruments, Santa Barbara, CA) operating in the tapping
regime mode using antimony doped silicon cantilever tips (POL-15, 130 do
250kHz , 48 N/m). Scanner was used with scan rates between 0.5 and 1 Hz. All
images are subjected to a first-order plane-fitting procedure to compensate for
sample tilt. The microstructure of polyurethane was investigated on micro
sections. These were prepared using a rotary microtom RM 2165 (Leica) with a
LN 21 cooling device working at -60C.
2.3.4 Image analysis
Binary images revealing hard domains agglomerates were produced via digital
processing of AFM images. The size and volume fractions of the hard domains
agglomerates in polyurethane and nanocomposites were determined by
measurements on their sections [15]. Linear covariance method was used for the
description of the distribution of particles. The image was transferred to
MicroMeter software and quantitative analysis was performed. The diameters
were randomly determined on 5 microphotographs.
2.3.5 Fourier transform infrared (FTIR) spectroscopy
Infrared spectra of PURs were collected using a FTIR spectrophotometer
(Thermo Electron Corporation model Nicolet 6700). Measurements were carried
out using attenuated total reflectance (ATR) technique. Each sample was
scanned 64 times at a resolution of 4 cm-1 over the frequency range of 4000400
cm-1. Analysis of FTIR data enabled to determine the carbonyl hydrogenbonding index (R). A straight baseline was drawn in the spectrum between 1780
WIT Transactions on Engineering Sciences, Vol 72, 2011 WIT Press
www.witpress.com, ISSN 1743-3533 (on-line)
Materials Characterisation V
381
cm-1 and 1640 cm-1 and the carbonyl stretching zone was corrected by
subtracting the baseline. To estimate the signal strengths, peak modeling of the
infrared active carbonyl bands was carried out using the Gaussian curve-fitting
method software OMNIC 7.3. The carbonyl absorption bands were deconvoluted
into component bands, the peak area of these bands was measured and carbonyl
hydrogen-bonding index R was calculated using Eq. (1) [16, 17]:
Abonded
A free
(1)
Moreover, the degree of phase separation (DPS) was obtained through Eq. (2):
DPS
R
R 1
(2)
Sample
E, storage modulus by frequency 5Hz at
37C, MPa
V, abrasion wear, dm3
24.6
2.8
32.6 30.8
1A
2A
3.4
40.6 41.6
3A
Materials Characterisation V
a)
Figure 2:
383
b)
c)
d)
e)
f)
50
1.
40
2.
30
20
10
0
0
0,5
1,5
2,5
SiO2 content, wt %
Figure 3:
Figure 4:
Sample
R
DPS
0
0.079
0.073
1
0.041
0.040
2
0.069
0.064
3
0.090
0.082
1A
0.071
0.,066
2A
0.105
0.095
3A
0.109
0.099
Materials Characterisation V
385
of hard domain agglomerates increases (Fig. 2d and f), which is the result of
reaction of NH2 groups on nanosilica surface and NCO groups in hard segments.
This reaction weakens chain mobility and creates favourable conditions for phase
separation in this nanocomposite group.
4 Conclusion
Within the framework of the work the influence of the quantity and modification
of nanosilica on the properties of polyurethanes was assessed; the polyurethanes
are intended to be used as short-term implants exposed to abrasion wear. The
addition of 0.5wt% of nanosilica caused a decrease in storage modulus and
abrasion wear of nanocomposites. Bigger amounts of nanosilica result in a
significant growth of storage modulus. The modification of nanosilica surface
with NH2 groups adds to the increase in the modulus and to a favourable fall in
the abrasion wear of the examined nanocomposites.
The joined qualitative and quantitative analysis of AFM pictures of
polyurethane structure and its nanocomposites as well as the analysis of phase
separation degree, performed on the basis of FTIR spectra, permitted to explain
the causes of changes in the properties brought about by the introduction of
nanofillers.
Acknowledgements
The authors are grateful to the Warsaw University of Technology for financial
support of this study. Thanks are also offered to Dr Maria Zielecka from the
Institute of Industrial Chemistry for the carrying out of nanosilica modifications.
References
[1]
[2]
[3]
[4]
[5]
[6]
[7]
[11]
[12]
[13]
[14]
[15]
[16]
[17]
[18]
[19]
[20]
[21]
[22]
[23]
[24]
[25]
[26]
[27]
Materials Characterisation V
387
Abstract
As part of this work oxydegradable polymer was fabricated and analyzed; it was
made from the waste following a 30 day exposure in Xenotest, simulating a
2-year exposure in atmospheric conditions and undergoing a triple injection
process. Similar examination was applied to composites containing 32 wt.% of
wood fabricated from this polymer. In the course of the examinations two
methods of degradation process evaluation of the materials were verified. It was
stated that the method utilizing the relationship between peak fields originating
from scissoring vibrations of the (-CH2-) group with a frequency of ca 1463cm-1
ensures more accurate results. The results of other examinations of the two
groups of materials permit us to state that the manufacturing of composites with
wood constitutes an interesting form of the utilization of oxybiodegradable
polymers.
Keywords: oxydegradable polyethylene, foil waste, recycling, wood.
1 Introduction
According to the data by Plastics Europe, European association of plastics
manufacturers, some 230 million tons of plastics were produced in 2009 in the
world (55 million tons in Europe). About 50% of the materials constitute
polymers intended for the packaging industry, of which 40% are polyolefines:
polyethylene (PE-LD, PE-HD, PE-LLD) and polypropylene (PP).
These polymers are made from petroleum-based synthetic polymers that do
not degrade in a landfill or in a compost-like environment. However, increased
use of synthetic packaging films has led to serious ecological problems. Several
WIT Transactions on Engineering Sciences, Vol 72, 2011 WIT Press
www.witpress.com, ISSN 1743-3533 (on-line)
doi:10.2495/MC110341
2 Experimental
2.1 Materials
Production waste, degradable polyethylene HDPE, containing 1 wt.% of prooxidant d2w, from Ecoplastic Poland, (OXY) was a kind of gift from Ecoplastic
Poland. Wood fibers Lignocel C 120 with particle size 70 150 m, from
J. Rettenmaier & Shne GmbH, Germany.
Materials Characterisation V
389
2.2 Sampling
Composite samples were fabricated in two stages. In the first stage, with the use
of T-45-25-22T-VS single-screw extruder with screw ratio L/D= 29 and with the
head for granulation able to simultaneously extrude six 5mm profiles, a granulate
was extruded containing 25 wt.% of wood flour. The extrusion process was
conducted at the temperatures of 150/160/170/175/175C (starting from the
charging hopper). After cooling the extruded rods were ground using a mill for
plastics. In the second stage wood flour was added to the granulate obtained
from the mixture in such quantity so that its total content in the composite should
reach 32% mass; next, the granulate of composite mixture (or oxybiodegradable
polymer) was placed in the bunker of DrBoy 22A injection moulding machine.
The regranulate and wood fibers as well as the obtained granulates were then
dried before each processing stage in DAC6 dryer at 80C for one hour. During
the injection of samples the cylinders temperature, starting from the bunker,
amounted to 155, 160, 170, 178 and 180C, injection pressure 800 105 Pa, and
clamp pressure of 800 105 Pa. Injection time was ca 2 s, clamp time 12 s,
cooling time 8 s; the total cycle time equaled ca 25 s. From all types of materials
samples A1 were fabricated in conformity with the norm EN ISO 527-2 using a
2-cavity mould cooled with water at 40C.
In order to verify the influence of injection process on the degradation process
of oxydegradable polymers and their composites a 3-time injection process was
carried out. Samples obtained in the first injection cycle were ground in an
industrial mill, dried and injected. This cycle was performed twice. A description
of the achieved materials is presented in Table 1.
Table 1:
Sample
1
2
3
1.32 2.32 3.32
Composites matrix
OXY OXY OXY OXY OXY OXY
Amount of wood
32
32
32
[wt.%]
Process multiplicity
1
2
3
1
2
3
2.3 Characterization
Resistance of the materials to accelerated ageing was determined on the basis of
analysis of pictures taken with the use of microscopic scanning, Charpy impact
tests as well as FTIR analysis of samples after irradiation. The examination was
carried out in conformity with PN-EN ISO 4892-1, PN-EN ISO 4892-2 and PNEN ISO 20105-A02:1996.
The samples underwent exposure in Xenotest Alpha High Energy equipped
with xenon lamp as radiation source. The examination was carried out for 720h
with samples being subjected to irradiation of 388,8MJ/m2 (within the scope of
300-400nm) equivalent to a 2-year exposure in natural conditions [19]. The
exposure in Xenotest is shown in Table 2.
Parameter
Filter
Luminous intensity E, W/m2
Work mode
Temperature control
Temperature in chamber C
Rain
Relative humidity, %
Phase time, min
Phase 1
Phase 2
Xenochrome 300
1503
1503
without
without revolution
revolution
in chamber
in chamber
353
303
no
yes
rain
503
102
18
SEM picture from the surface of examined samples dusted with gold was
achieved using an electronic scanning microscope Hitachi S-2600 with
accelerating voltage of 10 kV.
Infrared spectra of PURs were collected using a FTIR spectrophotometer
(Thermo Electron Corporation model Nicolet 6700). Measurements were carried
out using attenuated total reflectance (ATR) technique. Each sample was
scanned 64 times at a resolution of 4 cm-1 over the frequency range of 4000400
cm-1. Analysis of FTIR data enabled to determine the carbonyl index.
Charpy impact resistance using Resil 5,5 hammer by Ceast, wg PN-EN ISO 1792:2001 was defined for samples with notch, size 70 4 10 mm (cut from
samples and formed via injection).
Mechanical properties at static stretching in conformity with PN-EN ISO 5271:1998 and PN-EN ISO 527-2:1998 were examined using the strength machine
MTS Q/Test 10. Five 1A samples were analyzed from each type of composite.
The samples were stretched at the speed of 10 mm/min. The measurements were
recorded automatically using programme TestXpertII. Determined values:
strength in the area of plasticity (), strain at break ( ) and Young elasticity
modulus (E) of polymers and composites.
Absorption after water soaking was determined basing on the change in the
mass of 3 randomly selected samples from a given part of the material. The
examination was carried out in conformity with the technology specified in the
norm PN-EN 317:1999.
Materials Characterisation V
391
directly subjected to action of the xenon lamp (Fig. 1b). No such visible changes
were observed on the surface of composites (Fig. 2). Degradation of polymers
and composites results in the decrease in their impact resistance (Table 3).
Figure 1:
Figure 2:
Table 3:
Figure 3:
Basing on FTIR ATR spectra carbonyl index (COI) was calculated by two
methods. In the first one, proposed by Reddy and co-authors [16] and Corti et al.
[14], carbonyl index was determined as the ratio of absorbance of band 1716
cm-1 resulting from the vibrations of (C=O) carbonyl group and absorbance of
band 1468 cm-1 originating from scissoring vibrations of group (-CH2-).
COI
A1716
A1468
(1)
In the second method, proposed by Douminge and co-workers [18] and Stark
and co-workers [15] carbonyl index was determined as the ratio of absorbance
of band 1716 cm-1, resulting from the vibrations of (C=O) carbonyl group and
absorbance of band 2913cm-1, originating from the asymmetric stretching
vibrations of group (-CH2-).
COI
A1716
A2913
(2)
The fields of individual bands were determined with the use of OMNIC 7.3
programme. The analysis was carried out after correction of baseline, the
spectrum in the analyzed scope was resolved into component bands using the
Gaussian curve-fitting.
Materials Characterisation V
Table 4:
393
Sample
COI calculated acc. to formula (1)
COI calculated acc. to formula (2)
1
0.015
0.005
1.32
0.068
0.019
1.D
0.019
0.006
1.32D
0.041
0.012
In the results of the calculation of carbonyl index using the first method it was
stated that the index is three times higher than that calculated by the second
method according to formula (2) (Table 4). Therefore, it was agreed that the first
method will ensure a more accurate analysis of degradation changes in
oxybiodegradable polymers and composites with wood, fabricated from them.
To verify this statement an analysis was performed of carbonyl index of the
polymer and composite containing 32 wt.% of wood nanofiller after multiple
processing and using the first method (Table 5). Carbonyl index is used to
monitor the progress of oxidation process. An increase in COI was observed in
oxybiodegradable polymers and composites, which indicates that degradation
process in these materials occurs after each processing cycle. After consecutive
stages of processing the speed of COI changes in composite with wood
decreases, which means that the introduction of wood slows down the
degradation process. Also examined were the strength properties, impact
resistance and water absorption of the materials; the results are shown in Table 5.
Table 5:
Sample
COI calculated acc. to formula
(1)
U, impact resistance, kJ/m2
A, humidity content, %
E, elasticity modulus, MPa
, stress in the area of plasticity,
MPa
, strain at break, %
1
0.02
1.32
2
0.03 0.09
2.32
0.06
3
3.32
0.23 0.07
38.5
0.02
820
152
8.0
3.48
1740
19
36.7
0.02
210
105
9.0
3,43
410
26
31,2 11.0
0.08 3,12
205 390
80
27
450
4.4
280
4.9
120
5.7
4 Conclusion
Within the framework of the work a possibility of manufacturing practical
materials from oxybiodegradable polymers waste was assessed as well as their
composites with wood flour filler. The possibility of utilizing recycled
oxybiodegradable polymers requires the assessment of their degradation degree
WIT Transactions on Engineering Sciences, Vol 72, 2011 WIT Press
www.witpress.com, ISSN 1743-3533 (on-line)
Acknowledgements
The study has been financed by the National Research and Development Centre
within the framework of the project N R15 0023 06 / 2009, titled: Polymer
Composites with Biomass
The examinations in Xenotest were carried out by a team headed by Prof.
Ph.D. Eng. K. Czaja at the faculty of Chemistry, Opole University, Poland.
References
[1] Arnaud R, Dabin P, Lemaire J, Al-Malaika S, Chohan S, Coker M, et al.
Photooxidation and biodegradation of commercial photodegradable
polyethylenes. Polym Degrad Stab 1994; 46:211-24.
[2] Weiland M, Daro D, David C. Biodegradation of thermally oxidized
polyethylene. Polym Degrad Stab 1995;48:275-89
[3] Jakubowicz I. Evaluation of degradability of biodegradable polyethylene
(PE). Polym Degrad Stab 2003; 80:39-43.
[4] Magagula B. , Nhlapo N., Focke W.W, Mn2Al-LDH- and Co2Al-LDHstearate as photodegradants for LDPE film. Polym Degrad Stab 94 (2009)
947954.
[5] Roy P.K., Surekha P., Raman R., Rajagopal C., Investigating the role of
metal oxidation state on the degradation behaviour of LDPE, Polym Degrad
Stab 94 (2009) 10331039.
[6] Wiles DM, Scott G. Polyolefins with controlled environmental
degradability. Polym Degrad Stab 2006; 91:158192.
[7] Chiellini E, Corti A, Swift G. Biodegradation of thermally-oxidized,
fragmented low-density polyethylenes. Polym Degrad Stab 2003; 81:341
51.
Materials Characterisation V
395
[8] Hasan F, Shah AA, Hameed A, Ahmed S. Synergistic effect of photo and
chemical treatment on the rate of biodegradation of low density
polyethylene by Fusarium sp AF4. J Appl Polym Sci 2007; 105:146670.
[9] Albertsson A-C, Andersson SO, Karlsson S. The mechanism of
biodegradation of polyethylene. Polym Degrad Stab 1987; 18:7387.
[10] Albertsson A-C, Karlsson S. Three stages in degradation of polymers
polyethylene as a model substance. J Appl Polym Sci 1988; 35:1289302.
[11] Albertsson A-C, Barenstedt C, Karlsson S. Susceptibility of enhanced
environmentally degradable polyethylene to thermal and photo-oxidation.
Polym Degrad Stab 1992;37(2):16371.
[12] Sipinen AJ, Rutherford DR. A study of the oxidative degradation of
polyolefins. Proc Am Chem Soc 1992;67:1857.
[13] Hakkarainen M, Albertsson A-C. Environmental degradation of
polyethylene. Adv Polym Sci 2004; 169:17799.
[14] Corti A., Muniyasamy S., Vitali M., Syed H. Imam S.H., Chiellini E.
Oxidation and biodegradation of polyethylene films containing pro-oxidant
additives: Synergistic effects of sunlight exposure, thermal aging and
fungal biodegradation, Polym Degrad Stab 95(2010) 1106-1116.
[15] Stark N.M., Laurent M. Matuana L.M., Surface chemistry changes of
weathered HDPE/wood-flour composites studied by XPS and FTIR
spectroscopy, Polymer Degradation and Stability 86 (2004) 19.
[16] Reddy M.M., Deighton M., Gupta R.K., Bhattacharya S.N., Parthasarathy
R., Biodegradation of Oxo-Biodegradable Polyethylene, J Appl. Polym Sci,
111 (2009) 14261432.
[17] Jakubowicz I., Narahmadi N., Petersen H., Evaluation of the rate of abiotic
degradation of biodegradable polyethylene in various environments, Polym
Degrad Stab 91 (2006) 1556-1562.
[18] Douminge L., Mallarino S., Cohendoz S., Feaugas X., Bernard J., Extrinsic
fluorescence as a sensitive method for studying photo-degradation of high
density polyethylene part I, Current Applied Physics 10 (2010) 12111215.
[19] Ryszkowska J., Saasiska K., Kompozyty z folii oksybiodegradowalnej z
recyklingu napelniane drewnem, Polimery 55 (2010) 740-747.
Materials Characterisation V
397
Abstract
This paper reports the results of an investigation into the role of the slag
composition in recycling Oxygen Free High Conductivity (OFHC) copper
through a modified Electroslag Remelting Process (ESR) melting technique
Electroslag Crucible Melting (ESCM). Materials used for the slag were alumina
(Al2O3), cryolite (Na3AlF6), sodium fluoride (NaF), and fluorine (CaF2) at
different ratios. The results showed that in addition to purity of the slag and
graphite used in the electrode and crucible, the percentage of alumina component
in the slag composition was the main factor in the attainment of the purity. The
best consequence was achieved when the content of alumina was in the range of
2327%. Furthermore, in ternary compound systems with constant one
component, the influence of the weight ratio of the other components was
investigated, from a product purity viewpoint. Finally, in the optimized
condition, copper with 99.988% purity was achieved, yielding an electrical
conductivity of 100.1% IACS.
Keywords: OFHC copper, recycling processes, electro slag remelting, ESCM
process
1 Introduction
Recent developments in electronic industries and energy fields have pushed the
property requirements for oxygen free copper to the extreme. When recycling the
OFHC copper, the main problem is to reduce oxygen contamination because it
not only impairs the conductivity but also the mechanical properties of the
material, possibly leading to high scrap losses [14]. Oxygen free copper is used
in applications where the parts are going to be annealed in a hydrogen containing
Materials Characterisation V
399
metallic electrode. The electrode tip is heated by the hot slag and starts to melt
forming droplets of liquid metal, which disconnect from the electrode and sink
through the slag layer.
Figure 1:
The ESCM (Electro Slag Crucible Melting) method was first invented by
Borodin and co-workers in 1985 with inspiration from the ESR process [6, 16].
The main target was to design a system for remelting the scraps. The results of
analysis showed the fact that applying this method results in decreasing the
amount of Sn, Pb and S; although the Fe and Zn content may increase due to the
contamination of the charged scraps [2, 16]. The produced ingots are free of
impurities, voids, shrinkage and gas porosities. In addition, the problem of
hydrogen embrittlement due to annealing (at high temperatures about 400C) is
reduced substantially [6, 16].
In ESCM method, the consumable electrode in ESR (made of the charge
material) is replaced by a non-consumable graphite bar and a water-cooled
copper mold is superseded by a graphite one. Schematic illustration of ESCM
process is shown in Figure (2). This collection is set on a water-cooled base
plate. Scrap is fed in small pieces after complete melting of the slag. These
shavings are melted through passing the slag and form melt pool behind it.
Finally, the molten metal covered by slag is cast into a permanent graphite
mould [2].
Use of graphitic crucible and electrode makes it possible to reduce the oxygen
content of the melt remarkably. In addition, the high deoxidizing capacity of
graphite ceases the addition of deoxidizing agent, e.g. Lithium, copper phosphor,
etc. which usually decrease the electrical conductivity of copper. Besides, for the
reason of the very low solubility of carbon in copper (about 0.0025% in
WIT Transactions on Engineering Sciences, Vol 72, 2011 WIT Press
www.witpress.com, ISSN 1743-3533 (on-line)
Figure 2:
25C temperature), it can be used without the problem of carbon dissolution [6].
This method has several advantages such as simplicity of the equipment, the
lower cost and possibility of recycling scraps of OFHC copper besides nearly
uniform mechanical properties such as toughness in all directions [2, 6]. The
results of Prasad et al. experiments have shown that while using cryolite, the
content of oxygen decreases considerably. However, the harmful impurities such
as Si, Fe, and S attract into the melt to a substantial extent so the melt doesn't
have sufficient purity in this respect [17]. Therefore, in this research, it would be
attempted to reduce the amount of impurity elements especially Fe and Si via
altering slag composition. For this purpose, cryolite was replaced by mixtures of
sodium fluoride and alumina, two effective components in preventing oxygen
pick-up, with different ratios. An adequate fluidity of the slag as well as its
ability of providing the desired temperature was considered in determining the
slag composition.
1.1 Experimental procedure
An AC ESCM system with a capacity of 150 KVA, comprising a graphite
electrode of 50 mm diameter and a graphite crucible with internal diameter of
105 mm was used to do the tests. Copper scrap was in the form of pieces with 5
mm diameters and the lengths of 2-3 cm. These particles were dipped in a 10%
HCl solution for 20 min to remove the surface oxides and then dried in the air.
The slag combination consisted of alumina, fluorine, cryolite and sodium
fluoride in various ratios. The compositions are listed in table 1.
It can be seen from the table 1 that, various NaF to Al2O3 and CaF2 to Al2O3
ratios were investigated in constant CaF2 and NaF respectively. The effect of
slag recycling was also studied by reusing a certain slag in the next melting. 4 kg
of copper scrap and 2 kg of slag were used for each experiment. The slag
materials were preheated in a resistance furnace up to 600C for 2.5 h. The
process started with an arc between the graphite crucible and the electrode.
Materials Characterisation V
401
CaF2 (%)
Al2O3 (%)
Na3AlF6 (%)
NaF (%)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
10.5
30
30
30
40
10.5
30
30
30
23
35
40
33
35
35
37
4.5
40
35
30
15
4.5
12.5
40
35
30
17
20
15
27
15
25
23
85
15
85
30
35
40
30
-
25
-
30
35
40
60
45
45
40
25
40
40
Once the arc was stabilized, the slag was charged into the crucible. The heat
required for the process was generated due to Ohmic resistance of the slag pool
by the passage of electric current through it. When the slag was completely
melted, the copper pieces were gradually added to the molten slag and melted.
The molten slag covered the liquid copper and protected it from atmospheric
contamination during the process. The slag composition was selected properly,
so, passing the charge through, the oxide particles could be entrapped and the
clean liquid metal was collected at the bottom of the crucible. During crucible
melting, the melt and slag temperatures were measured using a thermometer. The
temperature of the melt was maintained constant at about 1500C by changing
the voltage of the power supply and/or changing the distance between the
electrode and the crucible bottom. After the copper was completely melted, the
power was kept constant at about 28 KV for 7 min in order to complete the
chemical reactions. Chemical analysis of the ingots was determined using an
analyzer, and the oxygen content was also specified by LECO (ASTM B 170).
The electrical conductivity was measured for some of the specimens by Ely
chemical plant (DIN 48200, 48201). Tensile tests were performed by a MTS
machine according to ASTM E 8M standard (D = 2.5 mm) and microstructural
observations were carried out after preparation of the samples, polishing, and
etching by a potassium dichromate solution for 10 sec.
2 Results
Chemical analysis of the primary OFHC scrap is presented in Table 2.
Table 3 shows the effect of the slag composition on the final analysis. The
amount of oxygen content and electrical conductivity for some of the samples
WIT Transactions on Engineering Sciences, Vol 72, 2011 WIT Press
www.witpress.com, ISSN 1743-3533 (on-line)
Fe S
Te
As Se Sb Cd Cr Pb Si Bi
Ag Zn Ni Co Mn Sn
<1 6.5 <1.2 <1 0.5 <1 3.5 <1 3.5 <1 <0.6 8.6 <3 1 <2 0.34 <1
Sample
No.
Fe
Si
Al
Other
1
2
3
4
5
2900
500
1500
200
360
2500
2500
2800
1700
730
260
110
30
20
20
750
1300
50
790
<1
700
1700
1000
100
360
15
-
230
320
30
<1
450
40
230
10
<1
<1
330
410
70
<1
700
290
420
30
<1
450
10
200
300
20
<1
100
11
200
570
10
<1
40
12
13
140
40
150
80
50
20
<1
<1
90
400
47
14
30
30
<1
130
42
15
16
17
11
10
32.6
101
78
<1
29.4
27.3
35.9
<1
<1
<1
10
-
33
12
18
Mn=40
Mg=700
Mn=40
Mg=800
Pb=230
Pb=60
Zn=240
Pb=60
Sn=40
Pb=50
Sn=20
Zn=40
Pb=60
Mn=20
Pb=60
Sn=20
Zn=80
Mn=20
Pb=40
Mn=10
Pb=30
-
Cu
purity
99.200
99.150
99.390
99.550
99.680
Conductivity
99.850
99.920
99.90
89.6
99.950
99.940
91.7
95.9
99.980
100
99.974
99.985
99.988
96.1
100
100.1
Materials Characterisation V
403
transferred from the slag into the first melt and then the recycled slag is of lower
contamination. Samples 8, 9 and 10 have the same production parameters as
samples 2, 3 and 4 respectively, but the higher purity of graphite used for
crucible and electrode; that leads to the higher purity of the final product.
Figure 3:
Figure 4:
3 Discussion
The aim of this research is the assessment of the capability of ESCM process in
recycling of precious OFHC scrap with acceptable product purity. Introduction
of oxygen into the melt leads to take place the oxide compounds. Since these
kinds of inclusions have fragile natures, the presence of such particles in the
microstructure, particularly on the grain boundaries results in lessening the
WIT Transactions on Engineering Sciences, Vol 72, 2011 WIT Press
www.witpress.com, ISSN 1743-3533 (on-line)
Figure 5:
Also, increasing silicon and sulfur content in the melt can be attributed to
reduction of their oxides in slag composition. These compositions are introduced
via fluorine and sodium fluoride [17].
Previous researches have mentioned oxygen content reduction using cryolite
as a slag component, but the results of this study determined that it's
accompanied by silicon, iron and sulfur increase in the melt. As Table 3
represents, the best slag composition (40% sodium fluoride, 23% alumina, 37%
fluorine) results in a product with a purity of 99.988% and electrical
conductivity of 100.1% IACS.
Materials Characterisation V
405
4 Conclusions
1.
2.
3.
4.
5.
6.
7.
References
[1] Y. Koshiba, T. Masui & N. Iida., "Mitsubishi Materials' High Performance
Oxygen Free Copper And High Performance Alloys", 2nd International
Conference on Processing Materials Properties, TMS, 2000.
[2] R.G. Baligidad, V.V.S. Prasad, G. Balachandran & V. Ramakrishna Rao.,
High Purity Copper Through Electro Slag Crucible Melting, Transactions
of Indian institute of metals,Vol.42, No.3, 1989, pp.339-342.
[3] D. Janicijevic., Modern Plant Produces Oxygen-Free Copper, Metal
Progress, Vol.79, 1961, pp.112-138.
[4] H.A. Blank., "Fabricating OFHC Copper ", Metal Industry, 4 June 1964,
pp.768-770.
[5] J. Crane & E. Shapiro., "High Conductivity High Temperature Copper
Alloy", US Patent No.3976477, 1974.
[6] A.I. Borodin & A.L. Evlevskii, Production of Copper Plates By
Electroslag Mould Casting, Advances in Special ElectroMetallurgy,
Vol.3, No.1, 1987, pp.8-9.
[7] B. Allison., "Copper- The Science and Technology of the Metal, Its Alloys
and
Compounds", Amer. Chem. Soc., Monograph Series, Reinhold
Publishing Corp., New York, 1954, p.244.
[8] V.K. Gupta, V.N. Madhav Rao & R.V. Tamhankar., Trans. Indian Inst.
Met., Vol.25,1972, pp.33-39.
[9] Anon., Metal Progressing 84 (1963)124.
[10] OFHC Copper, Production and Properties., TR No.125, International
Copper Development Council.
[11] H. Eckstein., "Working operations and their results for production of
oxygen-free Copper", NEUE HUTTE, Vol.3, No.1, 1958, pp.32-36.
WIT Transactions on Engineering Sciences, Vol 72, 2011 WIT Press
www.witpress.com, ISSN 1743-3533 (on-line)
Materials Characterisation V
407
Author Index
Abud I. ....................................... 77
Ahadi F. K. .............................. 397
Alhoseini S. H.......................... 397
Allison P. G. ............................ 287
Al-Qureshi H. A....................... 165
Alston S. M. ............................. 351
Antorrena G. ............................ 187
Athanasopoulou A. .......... 315, 327
Awietjan S. F. .......................... 263
Ayenk S.................................. 175
Bae G. H. ................................. 213
Barbosa C. ................................. 77
Bestley C. C. N. ....................... 351
Bigdeli Karimi M. ...................... 85
Bobrov M. .................................. 67
Boczkowska A. ........................ 263
Bouzid A.-H............................. 339
Brailovski V. ............................ 339
Brown S. G. R.......................... 351
Bruneton E. .............................. 155
Caminha I. ................................. 77
ern R. ................... 143, 199, 363
Chamis C. C. ............................ 109
Chandler M. Q. ........................ 287
Chicharro J. M. ........................ 275
Christou P. ............................... 131
Chupin S. ................................. 155
Colin J. ..................................... 307
Cummins T. K. ........................ 287
Cvahte P. .................................. 239
Dabek E. .................................... 67
Daniele V. .................................. 55
Darque-Ceretti E. ....................... 15
Dasan B.................................... 251
Dony A. ................................... 307
Dykman M. ................................ 27
Fiala L. ..................................... 143
Freire M. S. .............................. 187
Gabet O...................................... 27
Garland N. ................................. 95
Gonzlez-lvarez J.................. 187
Guagliardo P. ............................... 3
Gueit E. ...................................... 15
Hanaee K. ................................ 397
Herranz G. ............................... 275
Horgnies M. ......................... 15, 27
Hotza D.................................... 165
Huh H. ..................................... 213
Janssen R. ................................ 165
Karako C. ............................... 175
Karpuzova E. ............................. 67
Keide H...................................... 77
Kelekanjeri V. S. K. G. ............ 251
Ketabchi S. .............................. 397
Khan Z. ...................................... 95
Klemm A. J. ................................. 3
Ko V. .................................... 363
Koksal H. O. ............................ 175
Kollaros G. ...................... 315, 327
Kugler G. ................................. 239
Leite E. R. .................................. 41
Lombard S. ................................ 27
Mastromatteo F. ....................... 251
Michael A. ............................... 131
Mihulka J. ................................ 199
Montoro L. A. ............................ 41
Morales A. L. ........................... 275
Moser R. D. ............................. 287
Nascimento J. L. ........................ 77
Nieto A. J. ................................ 275
Northwood D. O. ....................... 85
Ould-Brahim N. ....................... 339
Ouni A. E. ................................ 307
Stoilov V. ................................... 85
Stroppa D. G. ............................. 41
Ramirez A. J. ............................. 41
Ros R. ..................................... 187
Rochais D. ............................... 155
Rodrguez G. P......................... 275
Roesler C. R. M. ........................ 77
Rushing T. S. ........................... 287
Ryszkowska J. ................. 377, 387
Vantsyan M. .............................. 67
Vzquez G. .............................. 187
Vejmelkov E. ......................... 143
Vishwanath T........................... 251
Vivet N. ................................... 155
Vborn J. ............................... 363
Saeed A. ..................................... 95
Sakurai H. ................................ 119
Saasiska K. ........................... 387
Samarin S. N. ............................... 3
Silva J. G. P. ............................ 165
Smith R. ..................................... 95
Sondhi S. K. ............................. 251
Song Y. P. ................................ 299
Taglieri G................................... 55
Tercelj M. ................................ 239
Tintillier P. ................................. 15
Turgay T. ................................. 175
As computer models become more reliable and able to represent more realistic
problems, detailed data is leading to the development of appropriate new types
of experiments.
Experimental measurements need to be conditioned to the requirements of
the computational models. Hence it is important that scientists working on
experiments communicate with researchers developing computer codes, as well
as those carrying out measurements on prototypes. The orderly and progressive
concurrent development of all these fields is essential for the progress of
engineering sciences.
This book resulted from the latest in a series of successful meetings that
provide a unique forum for the review of the latest work on the interaction
between computational methods and experimental measurements. The topics
include: Computational and Experimental Methods; Experimental and
Computational Analysis; Direct, Indirect and In-Situ Measurements; Detection
and Signal Processing; Data Processing; Fluid Flow; Heat Transfer and Thermal
Processes; Material Characterization; Structural and Stress Analysis; Industrial
Applications; Forest Fires.
WIT Transactions on Modelling and Simulation, Vol 48
ISBN: 978-1-84564-187-0
eISBN: 978-1-84564-364-5
Published 2009 / 672pp / 255.00
WIT Press is a major publisher of engineering research. The company prides itself on producing books by
leading researchers and scientists at the cutting edge of their specialities, thus enabling readers to remain at the
forefront of scientific developments. Our list presently includes monographs, edited volumes, books on disk,
and software in areas such as: Acoustics, Advanced Computing, Architecture and Structures, Biomedicine,
Boundary Elements, Earthquake Engineering, Environmental Engineering, Fluid Mechanics, Fracture Mechanics,
Heat Transfer, Marine and Offshore Engineering and Transport Engineering.
This book presents a general elastic and elastoplastic analysis method for the
treatment of two- and three-dimensional contact problems between two
deformable bodies undergoing small displacements with and without friction.
The authors approach uses the Boundary Element Method (BEM) and
Mathematical Programming (MP). This is applied to the contacting bodies
pressed together by a normal force which is sufficient to expand the size of the
contact area, mainly through plastic deformation, acted on subsequently by a
tangential force less than that necessary to cause overall sliding. The formulated
method described in this book is straightforward and applicable to practical
engineering problems.
Series: Topics in Engineering, Vol 45
ISBN: 1-85312-733-7
Published 2005 / 144pp / 55.00
This book presents a clear and well-organised treatment of the concepts behind
the development of mathematics and numerical techniques. The central topic
is numerical methods and the calculus of variations to physical problems. Based
on the authors course taught at many universities around the world, the text is
primarily intended for undergraduates in electrical, mechanical, chemical and
civil engineering, physics, applied mathematics and computer science. Many
sections are also directly relevant to graduate students in the mathematical and
physical sciences.
More than 100 solved problems and approximately 120 exercises are also
featured.
ISBN: 1-85312-891-0
Published 2005 / 408pp+CD / 165.00
There are large challenges facing the oil and gas industry, e.g. increasing
difficulty to find new oil and gas accumulations, a requirement for a clean
environment and reductions in CO2 emissions being voiced by public and
governments, requirement for long distance transport of natural gas from the
producers to the consumers and rapid changes in the price of hydrocarbons on
the world market. The decision makers need a wide background for their
decisions, and all aspects of the value chain are important in order to identify
the optimal field development solution.
The objective of this book is to treat in a consistent manner the key aspects
related to development of petroleum resources in order to allow an
interdisciplinary view of the topics involved in development projects and in
order to identify the most optimal development solution. The book is organized
to allow the reader to review the different topics individually so the book can
serve as a reference source for those interested in topics not familiar to
themselves.
This book is written as a reference book for training programs, especially
courses that are attended by students from many parts of the world, including
areas that are in early stages of hydrocarbon development as well as Norway
with its established offshore petroleum provinces. Although intended to be
useful in a wide range of contexts world wide, it has particular emphasis on
offshore developments, and is based to a large extent on insights from
Norwegian offshore projects.
ISBN: 978-1-84564-478-9
eISBN: 978-1-84564-479-6
Published 2010 / 288pp / 98.00
WITPress
Ashurst Lodge, Ashurst, Southampton,
SO40 7AA, UK.
Tel: 44 (0) 238 029 3223
Fax: 44 (0) 238 029 2853
E-Mail: witpress@witpress.com